This is one reason why there is so much confusion and misunderstanding concerning the Random Walk Thesis.
When I first studied this, many years ago, I thought Box and Jenkins were kidding when they analyzed the Dow as an ARMA(1,0), or a random walk. I figured they just put this in their textbook so they could distract investors away from a predictive model of the market while lining their pockets from knowing the true ARMA parameters. It was only after losing a bunch of money that I realized that they were not kidding. I should have realized this right away, since if they had been successful in the marketplace they would never be writing a textbook.
To be fair, though, the market is a first order autoregressive process with a trend, and it is that trend that makes money for long term investors. Furthermore, it is not surprising that the indices are random walks, because they follow the law of large numbers. That does not mean that individual stocks can not be modeled, at least for short periods of time. |
