Ok, im new, so be gentle. If any of you have read http://www.amazon.com/gp/product/02...103-1614200-5600649?s=books&v=glance&n=283155 which is a book on the mathematics of divorce & marriage, and basically this professor observes and videotapes a couples conversation for 60 minutes, and (this is from memory so the details are probably wrong), and figures out the units of measurement, and the degrees of variation for those units, and effectively has quantified the emotional reaction of this 60 minute conversation. big deal. well, this has allowed him a 95% accuracy rate in predicting divorce or not within 10 years (i think, cant remember exactly), and with a 15 minute observation period the accuracy drops to merely 90% - still incredibly good. im sure the times above are off, but the percentages are not, but anyway you get the idea. Basically, i'm wondering if anyone knows of efforts like this in relation to trading and finance. attempts to quantify and use as "indicators" things that are more difficult to quantify than the readily available numbers generates by financial statements, statistics, probabilities, etc. One example that comes to mind is monitoring an Associated Press feed (or other news feeds) and measuring the frequency of releases of certain financial instruments by scanning the text, quantifying that, backtesting and correlating (if possible) to the instruments and using it as some kind of public awareness or momentum indicator. ok, maybe not, but you get the idea. Any thoughts or flames?
Markets are discounting mechanisms; even if you were successful in your "quantifications", correct diagnoses do not necessarily yield profits.
I found tradeable patterns in certain kinds of releases from analysts. It worked iin 1999, haven't checked on it recently. I did a ton of research on key words in news releases, it was not real profitable to trade by itself, it still takes a human to match up the news story to the company and decide what is hot and what is not.