GMM with Weak Identification

James H. Stock, Jonathan H. Wright

Econometrica · 2000 · 836 citations · 46 references

Concepts

TL;DR

The paper develops asymptotic theory for GMM under weak identification and proposes confidence sets that are robust to this issue. The authors derive general weak‑identification results for GMM, specialize them to linear IV regression and CCAPM, and apply them to empirical CCAPM analyses. Numerical experiments confirm that weak identification causes conventional GMM to fail, and the new robust confidence sets lead to different empirical conclusions in CCAPM studies.

Abstract

This paper develops asymptotic distribution theory for GMM estimators and test statistics when some or all of the parameters are weakly identified. General results are obtained and are specialized to two important cases: linear instrumental variables regression and Euler equations estimation of the CCAPM. Numerical results for the CCAPM demonstrate that weak-identification asymptotics explains the breakdown of conventional GMM procedures documented in previous Monte Carlo studies. Confidence sets immune to weak identification are proposed. We use these results to inform an empirical investigation of various CCAPM specifications; the substantive conclusions reached differ from those obtained using conventional methods.

References

46