# Testing threshold cointegration in vector error-correction models

In Hansen and Seo's paper on Testing two regime threshold cointegration in VECM (J. Econometrics, 2002; 110:293), the authors proposed a test based on Lagrange Multiplier for testing treshold in model. However, suddenly the model denotes heteroskedasticity. Why the model denote heteroskedasticity?

-
Can you give us a link to the paper? There is no way anyone can helpou without seeing the paper. –  Michael Chernick Jul 23 '12 at 1:46
–  user12789 Jul 24 '12 at 9:55
please help me, my problem in section 3.1, why use heteroskedasticity? –  user12789 Jul 24 '12 at 9:56
Thanks I will take a look and will help if I can. But this may not be my area of expertise. –  Michael Chernick Jul 24 '12 at 10:31
i still waiting, thanks for help me –  user12789 Jul 25 '12 at 6:01
show 1 more comment