Navigation:  Menu Commands > Tools Menu > Variable Scenarios >

 

State Variable Scenarios

 

Print this Topic Previous pageReturn to chapter overviewNext page

Compute paths for the observed variables and state variables when a subset of the state variables are restricted to follow a certain path over a sub-sample. Parameter values are given by the initial or the posterior mode values, or a sample from the prior or the posterior distribution. In the event that a sample from the posterior is selected, the number of parameters in the sample is determined by the number of post burn-in posterior draws and the percentage use of posterior draws for impulse responses, variance decompositions and other functions of the parameters. These values are set in the posterior sampling frame on the Options tab.

The state scenario variables are linear combinations of the state variables. The potential scenario variables and their relations to the observed variables must be specified in the data construction file. Specifically, the field State can hold the sub-fields data, names, S and startperiod. The data sub-field is a matrix with the state scenario variable paths, names a cell array of string with the names of the variables, and S a matrix that maps the state variables into the state scenario variables. The startperiod sub-field is a vector with start year and start period for the scenario and is only used as a default value prior to having selected an alternative. For additional information, see the file  "DataConstFile.m" which is located in the directory "example\AnSchorfheide" directly below the base directory for YADA.

A subset of the scenario variables can be selected from the select state scenario variables function on the Actions menu. The selection of shocks and their connection with the scenario variables can also be selected from this menu.

The starting point of the scenario is a selectable period within the sample used to evaluate the log-likelihood function. That is, a period from the selected sample once the chosen training sample for the Kalman filter is skipped.

The user can choose between two shock selection methods:

User defined shocks and links to state scenario variables.
Optimally selected shocks.

The first method means that individual and user selected shocks are manipulated to achieve the scenario assumptions. The second method uses an optimization procedure, where the shocks are selected over the scenario sample such that, subject to the scenario assumptions, the squared deviation from the smooth estimates of the economic shocks based on the actual observed data is minimized.

In addition, both shock selection methods can be used for either the standard smooth estimates or the constrained smooth estimates.

The state variable scenarios can under both methods be performed for the variables in their original form or using the annualization data in the data construction file. In the latter case, only observed variables are displayed in annualized form.

 

Additional Information

A more detailed description about state variable scenarios can also be found in Section 11.11 of the YADA Manual.

NOTE: State variable scenarios are only available in the version of YADA that is exclusive to the NAWM team within the Directorate General Research of the European Central Bank and the Modelling Unit of the Monetary Policy Department of Sveriges Riksbank. The publicly available version of YADA does not include the code for this tool.

 

 


Page url: http://www.texlips.net/yada/help/index.html?state_variable_scenarios.htm