A Pretest to Differentiate Between Weak and Nearly-Weak Instrument Asymptotics

Abstract

We propose a pretest, bootstrap Kolmogorov-Smirnov test, to differentiate between weak and nearly-weak asymptotics. This is based on bootstrapping Wald Continuous Updating Estimator (CUE) based test. Since Wald CUE test has different limits under weak and nearly-weak cases this can be used in a pretest. We also conduct some simulations and show that some of the asset pricing models conform to nearly-weak asymptotics.