I think I made a similar suggestion many moons ago...and so I agree, it is a good suggestion. It would not be possible to create a program to find every error..or most errors..but it sure should be possible to build one that finds MAJOR errors...like if a stock splits 2/1 and it isn't recorded properly..or 3/1 , etc. Another possibility might be to check how many standard deviations in price..the recent price is from some moving aveage...and flag the big ones. Regarding missing data like dividends..it shouldn't be too hard to , say, first determine the average number of days between distributions (like most stocks would be quarterly)..and then flagging issues where x*these number of days have passed and there is no update. x might be 1.1 or 1.2 or something like that. Over time, such a program could be slowly built , tweaked and improved.
I was a data manager many moons ago and created such programs to improve the database i worked from so i know it is possible. It is lots of work initially..but long term, it saves you LOTS OF TIME and AGGRAVATION. --- In [email protected], daniel davis <[EMAIL PROTECTED]> wrote: > > How about creating a program to check the accuracy of the data base................. > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > > [Non-text portions of this message have been removed] >
