共 50 条
Long run recursive VAR models and QR decompositions
被引:4
|作者:
Hoffmann, M
[1
]
机构:
[1] Univ Southampton, Dept Econ, Southampton SO17 1BJ, Hants, England
关键词:
structural vector autoregression;
long run recursive VAR;
choleski decompositions;
QR decomposition;
economic fluctuations;
cointegration;
D O I:
10.1016/S0165-1765(01)00457-8
中图分类号:
F [经济];
学科分类号:
02 ;
摘要:
Long-run recursive identification schemes are very popular in the structural VAR literature. This note suggests a two-step procedure based on QR decompositions as a, solution algorithm for this type of identification problem. Our procedure will always deliver the exact solution and it is much easier to implement than a Newton-type iteration algorithm. It may therefore be very useful whenever quick and precise solutions of a long-run recursive scheme are required, e.g. in bootstrapping confidence intervals for impulse responses. (C) 2001 Elsevier Science B.V. All rights reserved.
引用
收藏
页码:15 / 20
页数:6
相关论文