Constructing Optimal Instruments by First-Stage Prediction Averaging

By G. Kuersteiner, Ryo Okui. 2010

78 citations recorded by the source corpus.

Subfields: Economics and Econometrics, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/da290f68fd680fab0d1837b305967673cbc35e53

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Statistics and Probability

Citations in the corpus

Weighted‐Average Least Squares (WALS): A Survey (2016, methodology): “Finally, Schomaker et al. (2010) examined the properties of various FMA estimators in the presence of missing data and Kuersteiner and Okui (2010) used the MMA estimator of Hansen (2007) to construct optimal instruments in the context of linear models with endogenous regressors.”
Model averaging, asymptotic risk, and regressor groups (2014, background): “In this case, write w = w1 = w∗ 1 and t = t2 so that (14) equals C∗ n (w) = w 2 (L1 − L2)− 2wσ(2)t = w(2)β̂ ′ X′Xβ̂ − 2wσ(2)t where the second line uses the substitution L1 − L2 = y′y − ê′ê = β̂ ′ X′Xβ̂.”
Generalized Least Squares Model Averaging (2016, unknown): “Similarly to Theorem 1 of Lewis and Reinsel (1985) or Lemma 3 of Berk (1974), we can write Â−1 m − A−1 m = Â−1 m (Âm − Am)A−1 m = [ A−1 m + ( Â−1 m − A−1 m )] (Âm − Am)A−1 m ,”
Model averaging by jackknife criterion in models with dependent data (2013, background): “It is worth noting that if Mn/n = O(1) (see Assumption 4 of Kuersteiner and Okui, 2010), then condition (13) is implied by the weaker condition ξ−2n Rn(w o m) = O(n −c̃) for any m, with c̃ being a positive constant.”
Select the Valid and Relevant Moments: An Information-Based LASSO for GMM with Many Moments (2013, methodology): “Kuersteiner and Okui (2010) recommend using the model averaging methods to approximate the optimal IV in the rst-stage regression.”