CMStatistics 2017: Start Registration
View Submission - CMStatistics
Title: Goodness-of-tit tests for GARCH models Authors:  Maria Dolores Jimenez-Gamero - Universidad de Sevilla (Spain) [presenting]
Simos Meintanis - University of Athens (Greece)
Sangyeol Lee - Seoul National University (Korea, South)
Abstract: Goodness-of-fit tests for correct specification of GARCH type models are proposed. This problem has been previously considered by other authors. Many of the proposed methods suffer from at least one of the following drawbacks: (i) being specific to a particular GARCH model, often of given (low) order, (ii) assume rather strong conditions (such as the existence of certain moments), and (iii) not being consistent under all fixed alternatives. On the contrary, our method is for arbitrary specification of the model; it will be seen to be consistent under arbitrary model deviations.