Navigation:  Tabs and Options > Bayesian VAR Tab >

 

Prior - Parameters on Lags

 

Print this Topic Previous pageReturn to chapter overviewNext page
Prior type: Lets you select the type of prior to use for the parameters on lagged endogenous variables. You can choose between three priors: (1) a Minnesota-style, (2) a normal conditional on the covariance matrix, and (3) a diffuse.
Prior mean for own first lag parameter on levels variables: Lets you select the prior mean of the parameter on the first own lag for all endogenous variables that are regarded as levels variables. Values between 0 and 0.95 are supported. The option requires that the prior type is either a Minnesota-style or a normal conditional on the covariance matrix prior.
Prior mean for own first lag parameter on first diff variables: Lets you select the prior mean of the parameter on the first own lag for all endogenous variables that are regarded as first difference variables. Values between 0 and 0.95 are supported. The option requires that the prior type is either a Minnesota-style or a normal conditional on the covariance matrix prior.
Overall tightness hyperparameter: Lets you select the overall tightness hyperparameter (λo) that influences the prior covariance matrix. Values between 0.05 and 5 are supported. The option requires that the prior type is either a Minnesota-style or a normal conditional on the covariance matrix prior.
Cross-equation tightness hyperparameter: Lets you select the cross-equation tightness hyperparameter (λc) that influences the prior covariance for parameters on variable j in equation i, where ij, when the Minnesota-style prior type is used. Values between 0.05 and 0.95 are supported.
Harmonic lag decay hyperparameter: Lets you select the lag decay hyperparameter (λh) that influences the prior covariance matrix exponentially. Values between 0.5 and 6 are supported. The option requires that the prior type is either a Minnesota-style or a normal conditional on the covariance matrix prior.

 

Additional Information

A more detailed description about how to set up the Bayesian VAR prior is found in Section 14.1 of the YADA Manual.

 

 


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