« Riding the Nobel Chariot | Main | Rob the Sort of Optimist »


Adithya Yaga

Dear Prof. Engle:
In your JEP paper on GARCH and ARCH (Garch 101; Journal of Economic Perspectives-Volume 15, Number 4-Fall 2001-Pages 157-168) you say the following:

"Data in which the variances of the error terms are not equal, in which the error terms may reasonably be expected to be larger for some points or ranges of the data than for others, are said to suffer from heteroskedasticity. The standard warning is that in the presence of heteroskedasticity, the regression coefficients for an ordinary least squares regression are still unbiased, but the standard errors and confidence intervals estimated by conventional procedures will be too narrow, giving a false sense of precision."

This seems to imply that the standard errors will *always* be too low resulting in high t-stats and low p-values. It is not obvious to me why this should always be the case. If you could provide me with some intuition or direct me to where I may find it, I would be greatful.

The comments to this entry are closed.