Ó¢ÎÄÎÄÏ×£ºAsymptotic normality of the QMLE in the level-effect ARCH model-ˮƽЧӦARCHÄ£ÐÍÖÐQMLEµÄ½¥½üÕý̬ÐÔ
Ó¢ÎÄÎÄÏ××÷ÕߣºChristian M. Dahl,Emma M. Iglesias
Ó¢ÎÄÎÄÏ×ÕªÒª£º
In this paper consistency and asymptotic normality of the quasi maximum like-lihood estimator in the level-effect ARCH model of Chan, Karolyi, Longstaff and Sanders (1992) is established. We consider explicitly the case where the parameters of the conditional heteroskedastic process are in the stationary region and discuss carefully how the results can be extended to the region where the conditional heteroskedastic process is nonstationary. The results illustrate that Jensen and Rahbek's (2004a,2004b) approach can be extended further than to traditional ARCH and GARCH models.
ÕªÒª±¾ÎĽ¨Á¢ÁËChan, Karolyi, LongstaffºÍSanders(1992)µÄˮƽЧӦARCHÄ£ÐÍÖÐÄ⼫´óËÆlihood¹À¼ÆÁ¿µÄÏàºÏÐԺͽ¥½üÕý̬ÐÔ¡£ÌÖÂÛÁËÌõ¼þÒì±ä¹ý³ÌµÄ²ÎÊýÔÚƽÎÈÇøÓòµÄÇé¿ö£¬²¢ÏêϸÌÖÂÛÁËÈçºÎ½«ËùµÃ½á¹ûÍƹ㵽Ìõ¼þÒì±ä¹ý³ÌµÄ·ÇƽÎÈÇøÓò¡£½á¹û±íÃ÷£¬Ó봫ͳµÄARCHºÍGARCHÄ£ÐÍÏà±È£¬JensenºÍRahbek (2004a,2004b)µÄ·½·¨¿ÉÒÔ½øÒ»²½Íƹ㡣