Hi Chuz, Enjoyed your post applying runs probability to the stock market. I suspect there are a great many situations where a sequence of similar events can be found which looks to be caused by an underlying process but which is in fact random. Take basketball, for example. During the course of a game, one team may score 7 or 8 baskets in a row. When these kinds of runs occur, fans often use the word "momentum" to describe them. In fact, of course, many of these runs are purely random. Of course, it is possible for the result to be non-random, as when a shooter gets hot. But runs will occur whether there is momentum or not. Momentum is one of the most excessively used words in sports IMO.
I found quite interesting your counter-intuitive calculation that, over the course of one year, the probability of getting at least 8 consecutive down days in the stock market is 61.4%. Momentum investors and others who look for short term trends in the market would be well advised to reflect on calculations of this type.
Just to add a modest footnote to your result, the probability we would see two or more runs of 8 consecutive down days or more is surprisingly high - .241. I arrived at this by finding the probability of exactly one run of 8 days or longer:
243(.0039606)(.996094^242) = .373.
Therefore, the probability of two or more runs is 1- (.373+.386) = .241.
Haven't had much time for SI recently, and am nearly a week behind in catching up on my favorite posters. Sorry if this response seems dated.
Geoff |
