Noting today the concurrent divergences in the moves in such highly correlated pairs as dollar yen, down, and dollar euro, up, or corn down below 3 and soybeans up and in general IBM way up above 90 versus the market up a gnat, I see many great applications of Goodman's paper "Movements and Comovements in Dependent Time Series."

What I like about the mathematical methods in the paper is that many of them depend on signs, and simple multiplications of one change versus another. The kind of thing that the back of envelope and pencil can really help up gain insight into, and that a little programming is very helpful in gaining insight and generating hypotheses from. Doc Castaldo here has a copy of the paper available in snail mail here, and if you contact him at doc<at>dailyspeculations<dot>com he will send.

One of things that inspires this is the revulsion I always feel when my work is compared to Jesse Livermore or layman Murphy who likes to study these things with visual displays that are concurrent randomness on top of randomness without any of the appropriate adjustments for comovement that Goodman recommends. Another, Astagfirullah, is a post of mine being compared to the work of a former lady Spec who likes to describe outside days, and the many serious articles from former SpecListers who speak favorably of the random performance of certain sports trendists.