Leave-two-out cross fitting bias corrected test for cross-section dependence
Title: |
Leave-two-out cross fitting bias corrected test for cross-section dependence |
DNr: |
SNIC 2022/22-1101 |
Project Type: |
SNIC Small Compute |
Principal Investigator: |
Simon Reese <simon.reese@nek.lu.se> |
Affiliation: |
Lunds universitet |
Duration: |
2022-11-11 – 2023-12-01 |
Classification: |
50201 |
Keywords: |
|
Abstract
This project aim at using a leave-two-out approach to bias-correct tests for cross-section correlation in residuals from linear panel data regression models. The leave-two-out approach addresses the problem of correlation induced by demeaning or defacoring data across cross-sections in a longitudinal/panel dataset.
The computational resources will be used for monte carlo simulations that are too computationally intense for my desktop PC at work.