Another Look at Bandwidth-free Inference: a Sample Splitting Approach

2026-05-30

AI Summary

This study addresses the pronounced size distortion exhibited by existing bandwidth-free tests under moderate parameter dimensionality and time series dependence when sample sizes are small to moderate. To overcome this limitation, the authors propose an inference framework that integrates sample splitting with self-normalization, effectively reducing high-dimensional testing problems to one-dimensional ones. Within this framework, they construct both $L_\infty$- and $L_2$-type test statistics applicable to a range of settings, including mean testing, autocorrelation assessment, linear regression hypotheses, and multivariate mean change-point detection. Notably, they derive the asymptotic null distributions of both statistics and establish their asymptotic independence under the null hypothesis in a setting where the parameter dimension grows with the sample size—a result achieved for the first time. Simulation studies demonstrate that the proposed method substantially mitigates size distortion and achieves more accurate empirical size and competitive power across various time series testing scenarios.

Abstract

The bandwidth-free tests/inferences for a multi-dimensional parameter have attracted considerable attention in econometrics and statistics literature. These tests can be conveniently implemented due to their tuning-parameter free nature and possess more accurate size as compared to the traditional HAC-based approaches, where consistent long run variance estimation was involved. However, when sample size is small/medium, these bandwidth-free tests exhibit large size distortion when both the dimension of the parameter and the magnitude of temporal dependence are moderate, making them unreliable to use in practice. In this paper, we propose a sample splitting based approach to reduce the dimension of the parameter to one for the subsequent bandwidth-free inference. Our SS-SN (sample splitting plus self-normalization) idea is broadly applicable to many testing problems for time series, including mean testing, testing for zero autocorrelation, linear hypotheses testing in a time series regression model and testing for a change point in multivariate mean. Specifically, we propose $L_{\infty}$-type and $L_2$-type SS-SN test statistics and derive their limiting distributions under both the null and alternatives and show their effectiveness in alleviating size distortion via simulations. As an important theoretical contribution, we obtain the limiting distributions for both SS-SN test statistics in the multivariate mean testing problem when the dimension is allowed to diverge as sample size grows to infinity. In addition we show the asymptotic independence of $L_{\infty}$-type and $L_2$-type SS-SN test statistics under the null in the growing dimensional setting.

Problem

• bandwidth-free inference
• size distortion
• sample splitting
• time series
• high-dimensional parameter

Innovation

• sample splitting
• self-normalization
• bandwidth-free inference
• high-dimensional time series
• asymptotic independence