Abstract
A vector time series model with long-memory dependence is introduced. It is assumed that, at each time point, the observations are equi-correlated. The model is based on a fractionally differenced autoregressive process (long-memory) adjoined to a Gaussian sequence with constant autocorrelation. The maximum likelihood estimators for the parameters in the model are derived and their asymptotic distributions are obtained.
Original language | English (US) |
---|---|
Pages (from-to) | 285-293 |
Number of pages | 9 |
Journal | Statistics and Probability Letters |
Volume | 31 |
Issue number | 4 |
DOIs | |
State | Published - Feb 1 1997 |
Keywords
- Asymptotic inference
- Long-memory dependence
- Maximum likelihood estimation
- Time series
ASJC Scopus subject areas
- Statistics and Probability
- Statistics, Probability and Uncertainty