interpreting TVECM, 2 regimes only ECT common to both regimes

[Betty Vel]

Hi everybody,
I am a PhD candidate an I am using the package from not so long and I have maybe a basic question. I do not know how to interpret the results from a TVECM model specification where I establish only ECT as common to both regimes.
Is ECT- the first regime ECT, and ECT+ the correction term of the second?
My command and results are below.
Many thanks
Beatriz

tprices1<-TVECM(prices1, nthresh=1,common=c("only_ECT"))
133 (5.3%) points of the grid lead to regimes with percentage of observations < trim and were not computed
> summary(tprices1)
#############
###Model TVECM
#############
Full sample size: 207 	End sample size: 205
Number of variables: 2 	Number of estimated parameters 10
AIC -3785.12 	BIC -3748.567 	SSR 0.04160168 

Cointegrating vector: (1, - 1.051274 )
[[1]]
                              ECT-               ECT+               Const                      it t -1            ger t -1        
Equation it  -0.0795(0.0002)*** -0.0203(0.3692)    -0.0009(0.1618)  0.1268(0.0661).    0.1008(0.0137)*
Equation ger 0.0881(0.0027)**   0.1538(2.3e-06)*** 0.0020(0.0291)*  0.2868(0.0033)** 0.6455(2.2e-23)***
                      
Threshold Values: -0.05875338
Percentage of Observations in each regime 13.7% 86.3%

[Matthieu S]

Hi Beatriz

The name of the parameter is admittedly a little bit confusing: what is common here are the lags, and what is regime specific is the speed of adjustment coefficient on the ECT. Hence "ECT-" stands for coefficient on deviations that are smaller then your threshold parameter, while ECT deviations above it.

Hope this helps, 

Matthieu

--
You received this message because you are subscribed to the Google Groups "tsdyn" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tsdyn+unsubscribe@googlegroups.com.
To post to this group, send email to ts...@googlegroups.com.
Visit this group at https://groups.google.com/group/tsdyn.
For more options, visit https://groups.google.com/d/optout.

[ahmed shafiq Joyo]

DiD you run sup-LM test before TVECM ?