Testing Regression Slope for Evidence of a Market Trend

Summary

The indicator uses the slope of a linear regression on closing prices as a measure of trend direction and intensity. It proposes testing whether the slope differs statistically from zero: a slope not distinguishable from zero is treated as a sideways market, while a statistically significant slope is classified as a trend. The calculation scales the slope by its estimated standard error and compares the resulting statistic with a threshold associated with a 95% normal confidence interval.

The threshold is presented as an approximation based on the normal distribution and a t-distribution with many degrees of freedom. The observation period is a parameter to optimize, but the document provides no empirical validation, market examples, or performance results. Its stated one-sided comparison also does not test for a significant negative slope symmetrically, and the approximation depends on statistical assumptions about the observations. The indicator is therefore a proposed regime filter, not evidence that a trading strategy using it will work.

Key ideas

• Regression slope is used to represent trend direction and intensity.
• The method treats a slope statistically indistinguishable from zero as evidence of sideways price action.
• It estimates a slope statistic using the observation period and an estimated standard error.
• The normal-based threshold is an approximation to a t-test decision rule.
• The document gives no empirical validation, and its stated comparison may not classify negative slopes symmetrically.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.