Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Dell Technologies Inc. -- Ignore unavailable to you. Want to Upgrade?

To: StockJock-e who wrote (65759)	9/16/1998 4:57:00 PM
From: Chuzzlewit	Read Replies (1) \| Respond to of 176387

Stockjock-e Re: Are you telling me that in all your years you have never used technical analysis?! Yes, that is precisely what I'm telling you. If TA were correct more often than not it would be picked up as statistically significant (meaning that it was statistically differentiable from the null case), but that is not the case in any study I have seen. Recently someone on the PSFT thread claimed that the then current price of the low 30's (around 33 as I recall) was the absolute bottom based on any number of technical indicators. I asked him whether he would repudiate TA if PSFT dropped lower. He said no. PSFT dropped lower in the next few days. Here is the really odd thing about TA. It wraps itself in the jargon of statistics, but misuses statistical concepts. For example, it generates a metric called "stochastics", but the meaning of the word is basically random and uncorrelated -- precisely the opposite of what TA claims by generating those stochastics. Here is something that I posted some time ago. One of the ways to test randomness of data is to perform runs analysis. That is, do random events occur in clusters, and if they do, do they conform to the expected patterns? Let me put it this way, if there are 250 trading days in the year, and if the probability if a stock going up or down on a particular day is 1/2, what is the probability of five or more consecutive down days? The probability of any sequence of five days being down is 0.5^5, or .03125. So the probability that any particular sequence will not have 5 down days is 0.96875. But, there are 246 sequences in a year. So the probability that there will be at least one run of five consecutive down days is 1-0.96875^246, or 0.999594. That's virtually a certainty. The probability of having 7 or more consecutive down days is a little over 85%. The probability of having at least 10 consecutive down days in a year decreases to about 21% -- not all that rare. This kind of probabilistic analysis puts to rest the notion of "momentum", which is the mainstay of TA. TTFN, CTC