Econometric meaning/interpretation of the intercept term (except the CAPM definition)

emilioalzamora1

Well-Known Member
Hi @David Harper CFA FRM,

I want to kick off the following discussion about the intercept term in an OLS regression (except our well known and familiar interpretation in a CAPM setting).

In case an OLS regression yields a positive or negative intercept which is significant what does this imply in detail?

My research so far has been quite sparse.

'Testing for Normality' (Thode) says 'the slope is an estimate of sigma and the intercept is an estimate of the mean'.

ok, so far so good but if we do have a positive/negative significant intercept what can we infer from this?

Gujarati writes that a negative intercept value very often does not have a viable economic interpretation. But what happens in the opposite case with a positive intercept term?

If the intercept term is negative where negative values are actually not possible in case where y-values represent certain weights, then the intercept does not have a meaning.

If the x-value = 0 has an no economic meaning (the air temperature) then the intercept term should have a meaning as well.
 
Last edited:
Top