Reid & Terry,
I think both of your charts here tell the L/R story. An obvious trend change occured 70 days ago so the old data points got thrown out and we needed new data. The next 27 days (until the Jan low) may be looked at as a transition period as the market "down-shifted" a couple of gears to find its new rhythm. Because this period is transitional, the data may be suspect for accurate L/R. Beginning with the Jan low, the slope is noticeably flatter.
So, I think your second chart Reid makes more sense today, until the lower line fails, then we zoom back out to the first chart.
The more I think about it, I know there is an L/R equation out there that helps you know when a set of data points becomes unuseful due to a pattern shift. For example, if the slope of the trendline changes significantly, as in our example above. For that matter, there are equations that give measure to the "validity" of an L/R model. Reid, I suppose the program in Stockcharts is using some of these equations.
Regression Analysis - those courses were pretty interesting in school, but I've never used the material in my work life, so it's all a distant fog to me now.
J.K. |
