You can make money trading, but it takes an ENORMOUS amount of labor. So it helps to scale it with a lot of capital and equipment.


But trading, for the most part, involves using the news and the ticker to predict what levels of buyers and sellers will show up nearby. A longer form of trading focuses on larger trends in the supply of and demand for investments of specific durations, without that much regard to their relative utility compared to competing investments of alternative durations, meaning without regard to the value of stocks.


So if we reduce your question to can someone, in the long run, do enough better than his competitors at predicting the actual business prospects of companies to make money, screening out all those other factors, I would say "Yes." I know I can do it in at least a limited subset of companies with businesses that I understand better than anybody else.


Problem is, 99% of businesses nobody really understands, so they can trade at crazy prices seemingly forever. Take Enron - nobody could have had any idea what was going on, and I wasn't even paying attention. If somebody knew, he apparently didn't have too much credibility or a following, and probably sold a 100-lot. But then again, somebody had to be selling on the way down.


Analyst earnings estimates are only accurate so long as earnings exhibit strong serial correlation, a slight trend, and a lot of positive steering from management. Every time earnings collapse, analysts are behind the curve by precisely the magnitude of the sequential collapse.


But if you are like a freak biophysicist, for example, you may be able to predict the prospects for pipeline drugs better than enybody else. Or if you run a children's day-care center in Harlem, you may command a good statistical sample for determining the prospects of certain video game manufacturers or clothing trends. Or if you run a cash register at a monster grocery store in the Midwest...


olra