Abstract
We study asymptotics of parameter estimates in conditional heteroscedastic models. The estimators considered are those obtained by minimizing certain functionals and those obtained by solving estimation equations. We establish consistency and derive asymptotic limit laws of the estimators. Condition under which the limit law is normal is studied. Further, bootstrap for these estimators is discussed. The limiting distribution of the estimators is not necessary always normal, and we present a real data example to illustrate this.
Original language | English (US) |
---|---|
Pages (from-to) | 1135-1153 |
Number of pages | 19 |
Journal | Communications in Statistics - Theory and Methods |
Volume | 32 |
Issue number | 6 |
DOIs | |
State | Published - Jun 2003 |
Bibliographical note
Funding Information:We would like to thank the Referee for his comments and suggestions, and for the reference Abraham and Thavaneswaran (1991).
Keywords
- Bootstrap
- Conditional heteroscedastic
- Estimating equation
- Minimum contrast estimation