This paper analysis a causality under the heteroskedasticity. That the error terms have a covariance in the time series analysis can not be estimated efficiently
If the covariance assumption have been broken, The OLS analysis method is overestimated ...
This paper analysis a causality under the heteroskedasticity. That the error terms have a covariance in the time series analysis can not be estimated efficiently
If the covariance assumption have been broken, The OLS analysis method is overestimated at the coefficient and the test value is incorrected in the significant level
For the causality, the unit root test decides stability level of the variable and the level is choosen between level instrument and difference instrument.
The result of the unit root test was proved that the first difference instrument had the stability for WPI, CPI, WAGE.
The optimal lags was adapted by the Akaike-Shwaltz lags method. The result of the Akaike-Shwaltz was appeared the optimal lags of 3-6.
To efficiently estimate was used the ARCH, GARCH models that the error terms have the heteroskedasticity.
In case of the Granger's causality, the WAGE was caused by WPI, CPI but In case of the optimal lags causality, the WAGE caused by WPI, the CPI was caused the wage.
The efficient estimator was estimated by the ARCH and GARCH models. The causality between WAGE and WPI in the ARCH and GARCH models was appeared that the WAGE was caused by the WPI.
*Professor of Economics, Dongguk University
**Ph. D Candidate, Dep. Of Economics, Dogguk University
***Ph. D Candidate, Dep. of Economics, Dogguk University