Navigation:  »No topics above this level«

 

Observed Variable Scenarios

 

Print this Topic Return to chapter overview

Compute paths for the observed variables when a subset of the 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 scenario variables are linear combinations of the observed variables (current and past) and an initial condition. The potential scenario variables and their relations to the observed variables must be specified in the data construction file in a similar manner as the conditional forecast assumptions are setup. Specifically, the field C is used instead of Z, and CBar instead of U. 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 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 observed 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 observed 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.

 

Additional Information

A more detailed description about observed variable scenarios can also be found in Section 11.10 of the YADA Manual.

NOTE: Observed variable scenarios are only available in the version of YADA that is exclusive to the Forecasting and Policy Modelling Division within the Directorate General Economics of the European Central Bank. The publicly available version of YADA does not include the code for this tool.

 

 


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