Quote:
Originally Posted by outspan
Some trading platforms have a stddev indicator, so I think it's absolutely feasible to estimate it (and predict it to some extent) for a certain time frame. If sddev moves back and forth, I can use a sub-model to take that in consideration too.
|
The problem with random walk and all of that stuff is it's based on the idea that market returns are normally distributed. They absolutely are not. I would strongly suggest reading Mandelbrot on the subject. He does a very good job of pointing out the flaws in classic financial/economic theory, and how that results in much larger than expected risks being taken, though I won't say he's solved the question of how to forecast price movement.