STATIONARITY AND PERSISTENCE IN THE GARCH(1,1) MODEL

被引:475
作者
NELSON, DB
机构
关键词
D O I
10.1017/S0266466600005296
中图分类号
F [经济];
学科分类号
02 ;
摘要
This paper establishes necessary and sufficient conditions for the stationarity and ergodicity of the GARCH(l.l) process. As a special case, it is shown that the IGARCH(1,1) process with no drift converges almost surely to zero, while IGARCH(1,1) with a positive drift is strictly stationary and ergodic. We examine the persistence of shocks to conditional variance in the GARCH(l.l) model, and show that whether these shocks “persist” or not depends crucially on the definition of persistence. We also develop necessary and sufficient conditions for the finiteness of absolute moments of any (including fractional) order. © 1990, Cambridge University Press. All rights reserved.
引用
收藏
页码:318 / 334
页数:17
相关论文
共 23 条
[1]   GENERALIZED AUTOREGRESSIVE CONDITIONAL HETEROSKEDASTICITY [J].
BOLLERSLEV, T .
JOURNAL OF ECONOMETRICS, 1986, 31 (03) :307-327
[2]  
Davis P. J., 1965, HDB MATH FUNCTIONS, P253
[3]  
DUDLEY RM, 1989, REAL ANAL PROBABILIT
[4]  
Engle R.F., 1986, ECONOMET REV, V5, P1, DOI [10.1080/07474938608800095, DOI 10.1080/07474938608800095]
[5]   AUTOREGRESSIVE CONDITIONAL HETEROSCEDASTICITY WITH ESTIMATES OF THE VARIANCE OF UNITED-KINGDOM INFLATION [J].
ENGLE, RF .
ECONOMETRICA, 1982, 50 (04) :987-1007
[6]  
ENGLE RF, 1986, ECONOMET REV, V5, P81, DOI DOI 10.1080/07474938608800101
[7]  
Geweke J., 1986, ECONOMET REV, P57, DOI DOI 10.1080/07474938608800097
[8]  
Gradshteyn I. S., 1980, TABLES INTEGRALS SER, V4th
[9]  
Hardy G. H., 1952, MATH GAZ
[10]  
Kolmogorov A. N., INTRO REAL ANAL