More than that - his analysis simply showed (in a fairly poor way, I might add) that there's serial correlation in daily returns. This is very very well known. However, autoregressive models can still have a random component. In fact, I can easily devise a random number generator that would give you the same results. Just because there's structure doesn't mean it's not a random process.
Jack Benny is walking down the street when a stick-up man pulls out a gun and says, "Your money or your life!" An extremely long silence follows. "Your money or your life!" the thug repeats. Finally Benny says, "I'm thinking!"
You are all as entertaining as it gets and I thank you for that. Now here is a question for you. Far more difficult than the mere warblings from great theoretical minds. "What has this thread got to do with making money from the markets" regards f9
The two values are 'up' and 'down'. According to the random walkers, those two things appear in a fixed ratio and the runs test automatically accounts for that. It doesn't matter if it's 50/50 or 70/30. The test provides very strong evidence that the run sequences are not randomly generated, regardless of the ratio of ups to downs.
OMG! That is truly funny! Congratulations! You score the maximum points in my book of ignorance! Wow, must be embarrassing as hell to know so little!
neither the academics nor the smart money consider a time series of the S&P 500 "random" in the sense you are using it and in fact thats not the terminology they use anymore (for good reason, they've progressed since the 70's) the closest definition to "random" as IID: independently and identically distributed. The S&P 500 is definitely NOT IID and the academics admit that. It roughly follows a geometric brownian process with drift. Again WITH DRIFT! not a coin toss. The big flaw in the brownian motion approximation is that volatility in markets is time varying. Take this a step further and add in other variable besides price and you'll find a mountain of good academic work supporting violations of EMH. As previous posters said. Your original post was naive.