|
Post by dg on Nov 4, 2008 12:48:16 GMT -5
Here are the conditions that set the signals:
|
|
|
Post by dg on Nov 4, 2008 13:03:39 GMT -5
Here is a sample of the analysis chart (which I usually don't need to present):
|
|
|
Post by dg on Nov 4, 2008 13:15:31 GMT -5
Here is part 1 of 2 of the stage definitions:
|
|
|
Post by dg on Nov 4, 2008 13:23:19 GMT -5
Here is part 2 of 2 of the stage definitions:
|
|
|
Post by dg on Nov 4, 2008 13:31:37 GMT -5
Here is the "normal" cycle of stages (perhaps better labeled as ideal):
|
|
|
Post by dg on Nov 12, 2008 21:17:25 GMT -5
The indicators are developed by analyzing the price chart in a very special way. Assume that you could only trade at close of each day. Assume that you can only go long. Then imagine every possible trade you could have placed over N market days. When you have done that, you have a large body of data (the trade potential map) that characterizes trading for the price history studied.
When performing this analysis, one finds that proper selection of N (determined by trial and error) is necessary. N too small is extremely jerky -- so much so that the actual changes of direction of any parameters that characterize the trade potential map give meaningless signals. And N too large, although very smooth, is counter productive by producing far too much lag in the signals.
When one has established the optimal N value, one can then set up statistical parameters that characterize the analysis data map. For N trade days, analysis produces N(N-1)/2 unique trade experiences using close values as trade values.
Although I won't divulge what my dg indicators represent, understand that some are based on simple statistical or probability theory and others are developed via complex interrelationships between combinations of simpler parameters. What I will tell you is that the relationships between dg1 and dg2 are of most importance to the longer term trader. The rest lend sophistication to those attempting to understand where the market is in its behavior cycle for anticipation purposes, so that trading guidance is possible outside of ideal trading intervals (which occur rarely).
The simplest strategy is to buy (or cover) at the inception of CHERRY stage and to sell (or short) at the beginning of SHORTING stage. Another simple strategy is to buy at the beginning of buy signal majority and to sell at the beginning of sell signal majority.
The major problem with the simple strategies is that they tend to release a good deal of the profit made from the trade prior to recognizing the prudent time to exit. To deal with this problem, I use the total model performance overview and the guidance I give is what I have learned to be the appropriate action during each stage of the trading cycle. By incorporating LEMON stage as a time for sell stops and ERLYBRD stage as a time for cover stops, trade performance is usually considerably enhanced. (additional guidance is gained from NOZZLE and ROOSTER situations as well -- but in the interest of keeping this discussion brief, I will leave it to the reader to discern said understanding from the detailed information presented prior to this post).
Understand that nothing can predict the future. When all is said and done, any model bases its understanding of where things are and what to expect on past data. As many an experienced trader knows, tomorrow is unknowable. When the unexpected happens, my model will be just as confused as anyone else. Thus, no matter where the trader takes his guidance, he/she had better have good discipline when dealing with trades that go sour -- for he who hesitates will truely be separated from a good deal of his money (sometimes in as little as a few hours).
dg
|
|