GEN: Basics/Defining Trend

[Bob Hunt <RHunt.066@xxxxxxxxxxxxxxxx>]

As a follow-up to previous discussions concerning trading basics, I
have a suggestion for those who have been frustrated in using moving
averages as a way to define trend.

The April TASC has an excellent article on using Linear Regression
tools to define trend and trend strength ("Identifying Market Trends"
by Jack Karczewski). I've only been working with these concepts for a
week, but they appear to show great promise. I especially like the
ability to define trend strength, something that's generally missing
when using moving averages.

I've attached an .ela of the indicators discussed in the article. I
downloaded this file from the Omega site.

One indicator that is discussed in the article, but is not included in
the .ela is Linear Regression Errors. From my limited experience in
working with it, it seems to be an excellent indicator of short term
turning points, in all time frames. I've included my interpretation of
the indicator from the article below. If others, better versed in
statistical analysis, interpret it a different way, please post your
suggestion.

Bob Hunt


{-----------------------------------------------
  	Study		: Linear Regression Errors	
	Last Edit	: 3/12/98
	Provided By	: Bob Hunt (c) 1998    RHunt.066@xxxxxxxxxxxxxxxx

	Notes:	Indicator constructed from the April '98 TASC article
"Identifying Market Trends"
-----------------------------------------------}

Input: Length(30);

Value1=LinearRegValue(C,Length,0);
If Value1<>0 then Value2=((C-Value1)/Value1)*100;

Plot1(Value2,"LinRegEr");
Plot2(0,"0");

-----------------------------------------------