Econometrica Vol. 55 No. 4 1987
Asymmetric Least Squares Estimation and Testing
Abstract
This paper considers estimation and testing using location measures for regression m odels that are based on an asymmetric least-squares criterion functio n. These estimators have properties that are analogous to regression quantiles, but are easier to calculate, as are the corresponding test statistics. Asymmetric least-squares tests of homoskedasticity and s ymmetry compare quite favorably with other tests of these hypotheses in terms of asymptotic relative efficiency. Consequently, asymmetric least-squares estimation provides a convenient and relatively efficie nt method of characterizing the conditional distributi on of a dependent variable given some regressors.
- DOI
- 10.2307/1911031
- Volume
- 55
- Issue
- 4
- Pages
- 819
- Sources
- bibtex:phds-export.bib crossref openalex