January 4, 2012


I'm probably data mining (in a not-intellectually-honest way, by just focusing on November, December and early-January price and volume data), but I found it interesting to take a deeper look into the volume data. The behavior of volume (remember, I'm graphing cumulative volume below) leading up to late-December is somewhat predictable (with an R-squared of nearly 1). Also, I'm not being precise or pure with my statistical analysis. I'm just navel-gazing.

But, isn't it interesting to observe the change in volume behavior over the last 10 days or so? The R-squared of this (red) data series is noticeably lower than the first (blue) data series

As it stands so far today, the volume of approximately 160,000 shares exceeds the slope (128,000) of the red data series' trend line.

No comments:

Post a Comment