Working Paper: NBER ID: w16793
Authors: Patrick M. Kline; Andres Santos
Abstract: We examine the higher order properties of the wild bootstrap (Wu, 1986) in a linear regression model with stochastic regressors. We find that the ability of the wild bootstrap to provide a higher order refinement is contingent upon whether the errors are mean independent of the regressors or merely uncorrelated. In the latter case, the wild bootstrap may fail to match some of the terms in an Edgeworth expansion of the full sample test statistic, potentially leading to only a partial refinement (Liu and Singh, 1987). To assess the practical implications of this result, we conduct a Monte Carlo study contrasting the performance of the wild bootstrap with the traditional nonparametric bootstrap.
Keywords: Wild Bootstrap; Misspecification; Higher Order Properties; Econometrics
JEL Codes: C12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| wild bootstrap (C69) | partial refinement (Y60) |
| misspecification (C50) | wild bootstrap performance (C69) |
| errors correlated with higher powers of regressors (C29) | wild bootstrap failure to match second-order terms (C69) |
| wild bootstrap limitations (C69) | discrepancies in approximate cumulants (C46) |
| degree of misspecification (C50) | wild bootstrap inference refinement (D80) |
| wild bootstrap consistency (C62) | wild bootstrap practical implications (C69) |
| wild bootstrap performance in some scenarios (C69) | comparison to nonparametric methods (C52) |
| wild bootstrap shortcomings under significant misspecification (C52) | performance of wild bootstrap (C59) |