Shades
I don't look a lot at volume, or rather it is rather secondary in my analysis. I look for volume tailing off in consolidations, for starts of new waves on high volume. Very traditional stuff just looking at the chart.
This is more like the chart I am looking at:
stockcharts.com[m,a]waclyyay[pb13!d34][vc60][iUh5,5!Li5,5]&pref=G
The weekly chart shows that the major upwave from 180 is done. So we need a correction.
So I look at stochs in particular. MACD shows something similar but I find it harder to interpret. I like to just look at moving averages themselves. I use a series where every second Fibbonacci number features: 5, 13, 34, 89 etc. I even construct Bollinger Bands based on them. Another powerful indicator that has some basis in econometric theory.
Then I look at E-Wave and there is a pretty perfect series of five waves up on a daily chart.
See there are two main dips in the stochs:
stockcharts.com[m,a]daclyyay[pb13!d34][vc60][iUh14,3!Li14,3]&pref=G
And within each wave there is a classic E-Wave breakdown.
I also consult with the people on the CF E-Wave thread.
Then there is my own model. This is based on a regression analysis over different time frames in moving windows and seeing how the regression coefficients change. With the experience I have and historical data I've tested it on (e.g. the 1987 crash) I know the more reliable signals. No-one else has likely come up with this model as they wouldn't have read the combination of econometric literature and the willingness to do unconventional and statistically "incorrect" things. I don't really know why it works but it does. There is a powerful sell signal on GOOG.
So I integrate all this info together and my read on the overall market in the US and Australia etc.
I use a bunch of indicators in other words that could make sense in time series theory or fractal geometry, and mostly just use the daily closing price data. I don't use stuff that obviously doesn't make any sense (Bradley turning points, pitchforks etc.) or is very esoteric, options volatility etc. - I do look at the QQV though.
On the overall market I look every day at the McClellan Summation and recently at the $BPNDX. It is a great contrarian indicator.
I haven't seen that movie Pi. Heard about it. I saw a crap movie set in Australia called "The Bank".
imdb.com
I won't spoil it for you. It's a bit unrealistic. I'm not at that level of accuracy :) LOL
With all this info I think it would be hard to train a model to integrate it. But if you know someone who does neural networks and fuzzy logic, maybe we can get together :)
Let's see what happens.
David |
