Markov switching TVTP result
Posted: Sun Mar 11, 2018 7:35 am
Hi there,
I estimated a MSTVTP model, but the result for the transition matrix part is very odd. The coefficients are very large, like 4.4, -30.2,-5.4 and 52.5. I understand the model is not really fit with the data but I just want to know how to explain these parameters are >1. If they are the probabilities, should they less than 1?
Thanks
I estimated a MSTVTP model, but the result for the transition matrix part is very odd. The coefficients are very large, like 4.4, -30.2,-5.4 and 52.5. I understand the model is not really fit with the data but I just want to know how to explain these parameters are >1. If they are the probabilities, should they less than 1?
Thanks