TAR model with cointegration

[vee]

Hi - I Would like to estimate a TAR model with 3 variables (GDP and two sets of price series). The regime is determined by one of these price series - 'positive' regime when the price change is positive, and negative otherwise (i.e. threshold is zero).

Grateful for an example on how to do this seperately for cointegrated variables, and variables that are not cointegrated.

Thanks!

[1ci]

Dear fee,

have you already estimated such a TAR model? Can you help me pls?

Thank you!