[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Data Quality



PureBytes Links

Trading Reference Links

Mark,
what they did say about FutureSource ?
Frank

Mark Brown wrote:

> Fellow Traders:
>
> Just to show I'm not down on everyone in this business, just most..
>
>     Futures' feature article for their September '99 issue (as introduced on
> the front cover) which I saw yesterday was about data accuracy among 10 of
> the most visible end-of-day data services.
>
>     I wanted to call your attention to the fact that even though CSI has not
> advertised with them for many months including the current September issue,
> they bravely named CSI as the best data vendor with the fewest errors over
> the period of study.  They published their article, written by Sheldon
> Knight, knowing that at least nine firms would not look good including Omega
> Research's historybank.com who had an error rate that was 392% higher than
> CSI's error rate.  Omega logged 106 errors over the
> three year study period of two commodities vs. CSI's 27 "errors" over the
> same period on the very same markets.
>
>     I thought you would like to know this because if you remember I asked
> John Sweeney why an advertisement suddenly appears from a given "system
> promoter" on his initial trading-method exposure in S&C magazine.  Futures
> apparently was doing a service to the public instead of to its advertisers
> and they told the story in an unbiased manner.
>
>     Dial Data, by the way, didn't do too well, Their error rate was 1,188%
> higher than CSI's and, in fact for the S&P500 they committed an error every
> third day over the three year period studied.  The closest challenger to CSI
> was MJK who had an error rate of only 244.4% higher than CSI.
>
>     Take a look at the article.  It included a simple chart showing the
> consequences of using a flawed service by comparing the trading results of a
> simple breakout system using the data supplied by CSI, Omega, and MJK,
> respectively.  There was a significant difference between the hindsight
> performance of the three data sources.
>
> Mark Brown
>
> PS  I am still having trouble with the promised automated downloader, other
> than that I appreciate CSI for their attention to detail..