The RATS Software Forum

Dear Tom,
I am looking for the code of the following paper to implement smoorh transition cointegration. Is there any written available code for this? I looked at regime switching course but just find univariate example for LSTAR model.
Regards,

Kapetanios, George, Yongcheol Shin, and Andy Snell. "Testing for cointegration in nonlinear smooth transition error correction models." Econometric Theory 22.2 (2006): 279-303.

The transition model itself is just a straightforward non-linear least squares. It's a two-step process of estimating the cointegrating vector by an Engle-Granger regression and then estimating a NLLS model on the EG residuals. Similarly, the test is just the KSS test applied to the residuals from the EG regression (i.e. running the difference on lagged level cubed, with augmenting differences).

Note well that the LM test is a test for a unit root---it is *not* a test of the transition model. (One of those cases of confusing the null and the alternative). A run-of-the-mill AR stationary process will also trigger a rejection of the unit root in the KSS procedure.