The paper discusses the issue of block length selection in the blockwise bootstrap approaches to cointegrating regressions in small samples. The advantages in handling serially correlated errors are emphasized for the blockwise bootstrap approaches. Block length rules proposed in the previous literature are discussed. Some related Monte Carlo results are given as well to study how the estimation results are affected by the selection of block length in small samples.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media
Moving block bootstrap;
Document Type: Research Article
Department of Decision Sciences and Managerial Economics The Chinese University of Hong Kong Shatin NT Hong Kong
Department of Economics University of Illinois at Urbana-Champaign 1206 South Sixth Street Champaign IL 61820 USA
November 1, 2003