On the Limiting Distribution of Sieve VAR($\infty$) Estimators in Small Samples

TL;DR

This paper examines the asymptotic distribution of sieve VAR estimators in small samples, highlighting issues with traditional limit-based inference and proposing bootstrap or standard asymptotics as remedies.

Contribution

It clarifies the limitations of existing sieve VAR inference methods and suggests practical solutions to improve confidence interval accuracy in small samples.

Findings

Limit arguments cause overly conservative confidence intervals.

Bootstrap resampling improves inference accuracy.

Standard asymptotics can be used as an alternative.

Abstract

When a finite order vector autoregressive model is fitted to VAR($\infty$) data the asymptotic distribution of statistics obtained via smooth functions of least-squares estimates requires care. L\"utkepohl and Poskitt (1991) provide a closed-form expression for the limiting distribution of (structural) impulse responses for sieve VAR models based on the Delta method. Yet, numerical simulations have shown that confidence intervals built in such way appear overly conservative. In this note I argue that these results stem naturally from the limit arguments used in L\"utkepohl and Poskitt (1991), that they manifest when sieve inference is improperly applied, and that they can be "remedied" by either using bootstrap resampling or, simply, by using standard (non-sieve) asymptotics.