SDS-SSO PID update, April 18, 2009 20:54 EST

Dear Service Subscriber, thank you again for subscribing to my online equity information service. the latest S&P 500 Index Arb Model analysis is available for download. you are able to browse the directory to access previous analysis releases. if you ever misplace your login, email me stating so using the original email address you provided when subscribing through paypal. filename: SDS-SSO_20090418-1925_PID-5_eval.xls note: ignore the missing data error message when opening the file. just click OK or continue until excel displays the data. this is an artifact introduced from converting my original file i created on the Mac to reading the data in a PC. Today's Commentary well i've finally got some data for you to evaluate using my latest PID optimization algorithm. the download file has 2 worksheets labeled as neutral and aggressive. both neutral and aggressive scenarios employed the same PID optimization workflow to re-balance SDS & SSO. the data is in a crude table form and not charts, but you are certainly welcome to making some plots of your own. the main difference with the aggressive scenario workflow is applying additional criteria to check for "bias" changes in the weightings. in between re-balances, one or the other ETF may "re-polarize", that is reverse rates of changes between the pair. "bias" changes are re-weightings but are not necessarily co-incident with re-balances. i plan to put together a graphical illustration of this so you understand the distinction. i'm very pleased with the results of this level of optimization of my model. here's the general backtest workflow i followed. 1) first, neutral weightings are computed as usual starting now further back in time. along with the computing the weightings, other parameters such as PID reference data are calculated. 2) second, PID reference data then needs to be "tuned" to determine optimum coefficients for P, I, D. the P, I, D data generally require tweaking by multiplying the reference data by an integer, a decimal or integer plus a decimal. this establishes the optimum behavior. 3) third, the tuned PID data is charted and used to determine points on the time line indicate re-balancing to establish the neutral (benchmark) scenario. this scenario is a prerequisite to establishing the aggressive scenario. 4) now another graph of data in addition to the PID information is used to generate the aggressive scenario. sometimes in between re-balancing the model needs to be re-weighted to reflect "bias" changes in the ETF pair behavior. 5) finally, the aggressive scenario is completed by applying the proper weights at all transitions, re-balances and re-weight points. excel is very handy, but it has gotten very cumbersome to use as a backtesting platform. at some point i may look into procuring MATLAB or MAPLESOFT to do all the number crunching and plotting. these applications integrate with excel so data you receive will always be in a easy to use format. note that subscribers will only need to apply the results of the workflow above to their own portfolios. subscribers normally will not need to worry about the day-to-day rigors of going through all these optimization steps. this update is issued to provide some background into aspects of what the product will entail. tomorrow i'll send out the most updated weightings data (specifically friday's EOD trading results) normally incorporated in an update. best regards, mike james Managing Member Equity Informatics, LLC phone:302-220-3864