<i>M</i>‐Estimation for regressions with integrated regressors and arma errors (Q4828182)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 2118744
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | <i>M</i>‐Estimation for regressions with integrated regressors and arma errors |
scientific article; zbMATH DE number 2118744 |
Statements
<i>M</i>‐Estimation for regressions with integrated regressors and arma errors (English)
0 references
24 November 2004
0 references
\(M\)-estimators
0 references
regressions with integrated regressors
0 references
ARMA errors
0 references
endogeneity
0 references
serial correlation
0 references
0 references
0.8137346506118774
0 references
0.8064695000648499
0 references
0.7872229218482971
0 references
The authors consider a time series model \(y_{t}=X'_{t}\beta+z_{t}\), \(t=1,\ldots,n\), where \(\{y_{t}\}_{t=1}^{n}\) is a set of observations, \(X_{t}=(x_{t,1},\ldots,x_{t,r})\) is a sequence of \(r\times1\) vectors of integrated regressors satisfying \(X_{t}=X_{t-1}+u_{t}\), and \(\beta\) is a vector of regression coefficients. The error process \(z_{t}\) is a sequence of unobservable stationary ARMA errors \(\phi(L)z_{t}=\theta(L)e_{t}\), where \[ \phi(L)=1-\phi_1L-\ldots-\phi_{p}L^{p}, \qquad \theta(L)=1-\theta_1L-\ldots-\theta_{q}L^{q}, \] \(L\) is the lag operator such that \(Lz_{t}=z_{t-1}\) and \(e_{t}\) is a sequence of independent errors with mean zero and variance \(\sigma^2\). The process \(u_{t}\) is a zero mean stationary process possibly correlated with \(e_{t}\). \(M\)-estimators are proposed and their limiting distributions are derived under a proper normalization. Monte-Carlo simulation compares the proposed parametric methods with the semi-parametric method of \textit{P. C. B. Phillips} [Econ. Theory 11, 912--951 (1995)].
0 references