BDS Test Interpretation
Posted: Thu Nov 30, 2017 8:37 pm
Dear Altruists, Good Morning.
I am doing my master thesis and using BDS test as one my tests for non-linearity in the stock return series.
However, I am confused with the BDS test results provided by Eviews. Because in many papers I saw that, the asymptotic normal distribution does not fit for GARCH (1,1) standardized residuals! But Eviews provides Z-statistics assuming the series follows asymptotic normal distribution. And, for GARCH (1,1) standardized residuals De Lima (1996) and Brock et. al (1991) suggested the critical values of Simulated Table in Brock et al. (1991, pp 274-279).
In Eviews website, the help section for the BDS test does not provide any indication on how can we test our hypothesis in case of GARCH (1,1) standardized residuals.
Looking for your kind support.
With regards
Tohidul Islam Khan
I am doing my master thesis and using BDS test as one my tests for non-linearity in the stock return series.
However, I am confused with the BDS test results provided by Eviews. Because in many papers I saw that, the asymptotic normal distribution does not fit for GARCH (1,1) standardized residuals! But Eviews provides Z-statistics assuming the series follows asymptotic normal distribution. And, for GARCH (1,1) standardized residuals De Lima (1996) and Brock et. al (1991) suggested the critical values of Simulated Table in Brock et al. (1991, pp 274-279).
In Eviews website, the help section for the BDS test does not provide any indication on how can we test our hypothesis in case of GARCH (1,1) standardized residuals.
Looking for your kind support.
With regards
Tohidul Islam Khan