I am writing an indicator to use COT report data in backtesting and encountered a problem: unfortunately the historical COT data files at http://www.cftc.gov/cftc/cftccotreports.htm do not contain report release dates, which can be important information when doing backtesting. Without this info I cannot know for sure when the data was updated (usually Friday, sometimes Monday).
I tried to use the Federal Holidays calendar at http://www.opm.gov/fedhol/ to calculate the Fridays which are exceptions, but the dates there don't make sense when comparing with the current COT calendar for 2007.
Emailed them. Got a reply saying there is no list of release dates and there are no plans to keep such a data table.
The calendar at http://www.cftc.gov/dea/deacotcalendar.htm goes back only to November 27, 2006.
Maybe someone can help me? I need only the dates when the releases were not on Friday, back to year 2000 would be nice.
When backtesting do you think it's safe to asume the data will be updated at 3:35 pm (5 minutes after official update time) or should I be more conservative? Last time I tested the data (.txt format) was updated on website ~1 minute after official time.
Thank you
I tried to use the Federal Holidays calendar at http://www.opm.gov/fedhol/ to calculate the Fridays which are exceptions, but the dates there don't make sense when comparing with the current COT calendar for 2007.
Emailed them. Got a reply saying there is no list of release dates and there are no plans to keep such a data table.
The calendar at http://www.cftc.gov/dea/deacotcalendar.htm goes back only to November 27, 2006.
Maybe someone can help me? I need only the dates when the releases were not on Friday, back to year 2000 would be nice.
When backtesting do you think it's safe to asume the data will be updated at 3:35 pm (5 minutes after official update time) or should I be more conservative? Last time I tested the data (.txt format) was updated on website ~1 minute after official time.
Thank you
).