A few weeks ago, the Financial Times of London announced an expansion in coverage of the commodity sector. This week they published an article cleverly titled “Commitment to accuracy of CFTC data in doubt,” discussing the impact of inaccurate large trader reporting by Newedge from March through May 2011 (for which the CFTC fined the broker $700,000).
FT reporter Gregory Meyer may have revealed a bit of bias and frustration, commenting that, “the data are maddeningly vague [without errors].” It is a natural human reaction to disparage what you do not understand. But the evidence Mr. Meyer offers up to support his premise is entirely unconnected:
“In a sign that even CFTC officials acknowledge how hard it will be to generate accurate data, they created what they call a “safe harbour for less than fully compliant reporting” until late March [2012].”
Mr. Meyer fails to note that the safe harbour only applies to the CFTCs new swap reporting regime, which is just being implemented in response to Dodd-Frank. This is hardly detrimental to our purposes since we refer only to standard futures and options contract figures.
Mr. Meyer’s case (and the reader) would have been better served by disclosing the full extent of the Newedge reporting problems. The FT article failed to mention that the CFTC noted in their order that Newedge’s “inability to file accurate and timely large trader reports” extended from at least 2009.
Correct or not, these kind of news articles generate doubts about the usability of COT data. I have never suffered from the illusion that government reports were without error. Error-free data would be an unexpected gift in any statistical endeavor.
In spite of whatever shortcomings may affect the COT reports, we have found the data quite valuable. Recent evidence of this is included in our year-end report (BR874, Dec. 26, 2011).
We do build in safeguards in our data collection and analysis procedures (which have allowed us to notify the CFTC of potential errors on numerous occasions). For instance–and this is only one example–although I typically display actual trader net positions in Bullish Review, I also examine percent net positions when preparing the newsletter. This tends to mitigate open interest reporting errors.
I am beginning my 25th year of writing Bullish Review and over this time have come to appreciate unflattering articles concerning the COT data. You will not see me sending a letter to the FT editor defending the CFTC or disputing the “maddeningly vague” appraisal. The more widely-held this perception, the greater our trading edge.