SI
SI
discoversearch

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 : THQ,Inc. (THQI)
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Raymond James Norris who wrote (3050)1/14/1998 3:34:00 PM
From: Sigmund  Read Replies (2) of 14266
 
>>Just as the post before yours said, MAs have no predictive power. Only signs of support and weaknesses in a stock's direction: whether it be short, intermediate, or long term.<<

Well.....if we all agree on this, I don't understand what the disagreement is about. You said that you couldn't disagree with me more and now you say that you also don't see predictive value in MA's.

That was my only point. I certainly believe very much in TA. Clearly Vic has done a good job of calling short term trading movements. Perhaps he uses MA's and will make me look completely ridiculous.

I am trained as a mathematician. I know something about statistics and how it is misused. Finding patterns with historical data and trying to use this to predict the future is a tricky business. The entire basis of statistics is the test of a single repeat single hypothesis. Most people including many practicioners are not aware of this requirement.

When you test multiple hypotheses (e.g. multiple MA periods) you violate this principle on which statistics is based and your ability to use traditional statistics to produce confidence intervals for predictive purposes is destroyed. I imagine that if you are testing a small number of hypotheses the reduction in reliability is not that great, but when the number of hypotheses is expanded so that you are essentially just trying to find a relationship that fits, you no longer can use traditional statistics for predictive purposes. It is always possible to find a relationship that fits well with any set of data including random numbers. But that doesn't imply anything about the predictive ability of the relationship.
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  

HomeMailHotSubjectMarksPeopleMarksKeepersSettings
Terms Of UseContact UsCopyright/IP PolicyPrivacy PolicyAbout UsFAQAdvertise on SI
© 2025 Knight Sac Media.  Data provided by Twelve Data, Alpha Vantage, and CityFALCON News