vars (version 1.1-9)

vec2var: Transform a VECM to VAR in levels

Description

An object of formal class 'ca.jo' is transformed to a VAR in level presentation.

Usage

vec2var(z, r = 1)

Arguments

z

An object of class 'ca.jo' generated by function ca.jo() in package 'urca'.

r

The cointegration rank (default is r=1).

Value

• A list with class attribute vec2var holding the following elements:
• deterministic
  The matrix of deterministic coefficients.
• A
  A list with matrix object(s) containing the coefficients for the lagged endogenous variables.
• p
  The lag-order of the estimated VAR-process.
• K
  The count of endogenous variables.
• y
  A dataframe with the endogenous variables in levels.
• obs
  An integer signifying the count of used observations.
• totobs
  An integer signifying the total number of observations, i.e including observations taken as starting values..
• call
  The call to vec2var.
• vecm
  The supplied object z of formal class ca.jo.
• datamat
  A dataframe with the used dataset.
• resid
  A matrix with the residuals from the empirical VAR(p).
• r
  Intefer, the assigned co-integration rank from the call.

encoding

latin1

concept

• VAR
• Vector autoregressive model
• VAR
• VECM

Details

This function enables the user to transform a vector-error-correction model (VECM) into a level-VAR form. The rank of the matrix $\bold{\Pi}$ has to be submitted, i.e. how many cointegration relationships have been determined according to the outcome of ca.jo().

References

Hamilton, J. (1994), Time Series Analysis, Princeton University Press, Princeton. L�tkepohl, H. (2006), New Introduction to Multiple Time Series Analysis, Springer, New York.

See Also

ca.jo, predict, irf, fevd, Phi, Psi, normality, arch, serial, logLik, plot

Examples

library(urca)
data(finland)
sjf <- finland
sjf.vecm <- ca.jo(sjf, constant = FALSE, type = "eigen", K = 2,
spec = "longrun", season = 4, ctable = "A2")
vec2var(sjf.vecm, r = 2)