In the ARCitation[1] time series model (0 ≤ ϵ < 1)
If L denotes the likelihood function of y
1, y
2,…,yn
(conditional to , Model 2 of Vinod and Shenton Citation[2]) and z
(i) = (y
[i] − ϵy
[i]−1)/σ(1 ≤ i ≤ n) are the ordered variates (for a given ϵ), then the likelihood equations are (Tiku et al. Citation[1])
Following Islam et al. Citation[3], we do not simplify Citation[3] but linearize z (i) −1, namely,
Modified likelihood equations are obtained by incorporating Citation[4] in Citation2-3. The solutions are the following MML estimators:
Since β
i
> 0 (1 ≤ i ≤ n), is always real and positive. From the results given in Akkaya and Tiku Citation[4], it follows that
are asymptotically fully efficient, i.e., they are unbiased and their variance–covariance matrix is I
−1, where I is the Fisher information matrix which exists for all k > 2.
Efficiency: To compare the efficiencies of with those of
, in situations where the linear estimator
is always positive, e.g., when εi have one of the following distributions
Table 1. Means and Variances of the Estimators; ϵ = 0.5, σ = 1
It may be noted that the MML estimators above are enormously more efficient than the least square estimators Citation[1], Citation3-4.
Unknown location: In certain situations one may work with the model
Computations: The estimators above are computed exactly along the same lines as in Akkaya and Tiku Citation[4], i.e., are computed first in two iterations;
is then computed from Equation9 in a single iteration.
Acknowledgments
REFERENCES
- Tiku , M. L. , Wong , W. K. and Bian , G. 1999 . Time Series Models With Asymmetric Innovations . Commun. Stat.-Theory Meth. , 28 : 1131 – 1160 .
- Vinod , H. D. and Shenton , L. R. 1996 . Exact Moments for Autoregressive and Random Walk Models for a Zero or Stationary Initial Values . Econometric Theory , 12 : 481 – 499 .
- Islam , M. Q. , Tiku , M. L. and Yildirim , F. 2001 . Non-normal Regression: Part I, Skew Distributions . Commun. Stat.-Theory Meth. , 30 ( 6 ) to appear
- Akkaya , A. D. and Tiku , M. L. 2001 . Estimating Parameters in Autoregressive Models in Non-normal Situations: Asymmetric Innovations . Commun. Stat.-Theory Meth. , 30 ( 3 ) to appear