How does heteroskedasticity affect the theoretical properties of the OLS estimator?
Consequences of heteroskedasticity
1) OLS estimators are still unbiased & linear. But they are no longer BLUE, in both small & large samples .
Estimators don't have minimum variance & hence they are no longer efficient.
2) usual formulas of variance of OLS estimators are biased.& Overestimate true variance & standard deviation
3) this leads to inefficient forecasting.
Usual confidence intervals & hypothesis tests based on t& F distribution are unreliable, leads to wrong conclusions.confidence intervals are likely to be larger.
Get Answers For Free
Most questions answered within 1 hours.