SDS-SSO Model Update - June 5, 2009

Dear Blogger, thank you! again for subscribing to the S&P 500 long/short ETF Model and welcome to new members who just joined and new list subscribers. today's analysis is available for download: SDS-vs-SSO-20090605_subscriber.xls.zip - (or feel free to browse the directory.) ignore the missing data error message that may pop-up when opening up the file. Excel for Windows looks for metadata that Excel for Mac doesn't generate. if you ever misplace your login, send me a message using the email you originally provided when subscribing through paypal. visit the blog for an archive of all subscriber updates and alerts. the archive is search-able and comments can be posted by everybody. Today's Commentary the keyword in this commentary is geodesic. what's a geodesic? given by houghton mifflin a geodesic is "shortest line between two points on any mathematically defined surface." if the surface is a plane, the geodesic between two pts on that plan would be a straight line. if the surface has curvature, then the geodesic between two pts on the surface a curved line. what does this lesson in non-euclidean geometry have to do with SDS & SSO? follow along. because SDS & SSO are tied to the same index but move proportionally equal in opposite directions (approximately), we would not be wrong to expect there is a simple mathematical relationship between the rates of change between the 2 ETFs. unfortunately, the relationship exists but it is not simple because of tracking errors and compounding of those errors over time. computation of the neutral weights is the core of the solution to this problem. but there's a lot more to learn from these weights than meets the eye. consider the square of each weight as the magnitude of a vector. then hold each vector at 90 degrees to each other, add them together and resultant is vector c. the magnitude of vector c is the hypotenuse of the right triangle and theta is the defined by arc tangent (neutral wt of sds)^2/(neutral wt of sso)^2. if you plot the ||vector c|| versus theta, you have the geodesic or shortest path along a curved surface. i have included this plot in the spreadsheet. for every value of theta there is a unique value of vector c but not visa versa. such a plot is called a functional and is consistent with properties of a geodesic. each dot in the plot represents the results at the end of the day. it would not be correct to assume the adjacent dot corresponds to the results of the previous day or the day after in questions. the system may hop past previous values but only along the geodesic. because we are working with 2 variables, the geodesic should be smooth and symmetric. if the system ever breaks off symmetry, we can no longer assume we have neutral weights for the 2 ETFs. i will illustrate more points regarding vector c and it's relationship to SDS & SSO all throughout my commentaries this week. best regards, mike james Managing Member Equity Informatics, LLC phone:302-220-3864