You wrote...
>I wonder how someone could capture all that into some kind of moving
>average or percentage LOL
Not likely, but it is fathomable that "all that" (qualitative aspects of a stock's trading environment as well as fundamentals and traditional TA) could be captured by an artificially intelligent engine driven by a combination of expert systems and neural networks. Everything, down to the part of town in which a company has its offices, could be weighed against a database of "experience" compiled by hundreds or even thousands of "experts" (investors). An on-line agent would scan the internet, 24/7, for any discussion on a given stock. Taking a number of variables into account (stock exchange, trading history, liquidity, tendency towards manipulation), the agent would capture all relevant messages and, using a natural language parser, glean any useful DD, measure the ratio of hype to bashing, try to determine what kind of effect all of this could have on the stock price, and apply it to its analysis of the stock (which would initially be a "synopsis" based on non-controversial fundamentals and available news releases).
I have been working on the preliminary design of such a system for well over a year, and as soon as I get some free time to code... :-)
Michael Bendner |
