Page 1 of 1

Back-adjusted Data in CSI

Posted: Tue Jan 27, 2009 2:32 pm
by zumo
Hi All, I gather there are lot of posts here regarding back-adjusted data available from CSI, of which i am also a subscriber, but they haven't answered the questions i have regarding this apologies if the questions somehow are repetive from the postings..

I am currently using the back-adjusted portfolio that Blox provides for usage with CSI. I have some specific questions regarding what some of the settings entail..the settings for the contract (using t-note futures as an example here) and some of my own calculations are posted below.

Q1. As my calculations confirm, the continuous data series being created is back-adjusted and not forward-adjusted. What then is the "Generate Forward" check box thats enabled.

Q2. In my portfolio setting, the "Use only the C++ Back Adjusted" is checked on..from what i understand, the C++ back-adjusted performs forward-adjustment versus the Fortran Back-adjuster, which performs back-adjustment (CSI seems to use the term back-adjusting in its settings loosely wherein back-adjusting could imply forward or back-adjusting)..this is leading to the same confusion as in Q1 above

With regards to my own calculation, u can see that on 11/25th, the march 09 contract had higher open interest than the then-current dec 08 contract..and this set 11/25th as the roll-forward date. The roll adjustment to use was 1.9375, the difference between the close of these 2 contracts on 25th. This roll adjustemnt was used to create the adjusted prices for dec 08 contract for 11/24th and earlier..

The above adjusted price for dec 08 contract for 11/24th and earlier and the actual march 09 prices for 11/25th onwards fully reconciles with the back-adjusted data from csi using the portfolio provided by Trading Blox.

[Back-adjusted closes from CSI are Highligheted in Yellow and those calculated by myself are in Highlighted in Green]

Hope the above manual verification i performed helps others and hopefully somebody can help answer my questions.