Estimating the Speed of Adjustment in Partial Adjustment Models
比较了两种渐近等价的估计量在部分调整模型中的表现，发现模型设定比估计量选择对调整速度影响更大，对库存研究有参考价值。
We reexamine a recent controversy in the inventory literature regarding the impact of estimator choice on estimated speeds of adjustment in partial adjustment models. We compare the properties of two asymptotically equivalent estimators, the residual-adjusted Aitken (RAA) and nonlinear least squares (NLS) estimators. Our results indicate that (a) the RAA estimator is very sensitive to the quality of the instrument used in first-round estimation, and (b) model mis-specification may cause substantial differences in the finite-sample distributions of the two estimators. An empirical analysis of inventory equations suggests that adjustment speeds are more sensitive to model specification than to choice of estimator.