jmaih
commented
3 months ago Dear Gabriel,
If you've set up your model properly, all you need to do is filtering.
e.g. f=filter(m)
where m is a parameterized model object containing all the relevant information for filtration. or alternatively
f=filter(m,varargin)
where you pass the additional options you need.
Once you filter, f will be a structure containing information about the filtered variables, the updated variables and the smoothed variables. Some of those variables are going to be the probabilities of being in the different regimes.
Estimation is just going to be a change in parameters but the same principles will apply.
In terms of restrictions on the parameters, check out an example in \examples\VariousModels\TaoZha\LWZ_Econometrica2013\master_switch
Cheers,
J.
On Sun, May 26, 2024 at 8:36 PM GabrielPestana81 @.***> wrote:
Dear Dr. Maih,
I have a very simple model, and two calibrations that lead to two different regimes. In one calibration, I have a passive fiscal policy, which is a standard model. In the other calibration, I have an active fiscal policy. They differ only by two parameters, which are the monetary policy reaction to inflation ($\alpha_\pi$) and the fiscal policy reaction to the debt size ($\alpha_b$).
Given a set of data, could please point me out how could I use the toolbox to obtain estimates of probability of being in each regime in sample?
Later, I would also like to estimate the parameters in the model. However I would have to set as priors the relevant areas for the parameters that changes between regimes. That is, in the passive fiscal policy the prior must ensure $\alpha_\pi \geq 1$ and $\alphab > 1/\beta - 1$, and with active fiscal policy $\alpha\pi < 1$ and $0 < \alpha < 1/\beta
- 1$.
Thanks in advance !
— Reply to this email directly, view it on GitHub https://github.com/jmaih/RISE_toolbox/issues/185, or unsubscribe https://github.com/notifications/unsubscribe-auth/AATKBT7XY6RPBLJH2THPRPTZEITS5AVCNFSM6AAAAABIKAIEK6VHI2DSMVQWIX3LMV43ASLTON2WKOZSGMYTOOBYGA4TENQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>
GabrielPestana81
Dear Dr. Maih,
I have a very simple model, and two calibrations that lead to two different regimes. In one calibration, I have a passive fiscal policy, which is a standard model. In the other calibration, I have an active fiscal policy. They differ only by two parameters, which are the monetary policy reaction to inflation ($\alpha_\pi$) and the fiscal policy reaction to the debt size ($\alpha_b$).
Given a set of data, could please point me out how could I use the toolbox to obtain estimates of probability of being in each regime in sample?
Later, I would also like to estimate the parameters in the model. However I would have to set as priors the relevant areas for the parameters that changes between regimes. That is, in the passive fiscal policy the prior must ensure $\alpha_\pi \geq 1$ and $\alphab > 1/\beta - 1$, and with active fiscal policy $\alpha\pi < 1$ and $0 < \alpha < 1/\beta - 1$.
Thanks in advance !