Scribd
Upload
76 views12 pages

VAR Slides

Uploaded by

moira142560
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
76 views12 pages

VAR Slides

Uploaded by

moira142560
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

Vector Autoregression (VAR)

Dr. Elya Nabila Abdul Bahri

1
Overview of VAR Model

• VAR models in economics were made popular by Sims (1980).


• natural extension of the univariate autoregressive model to dynamic
multivariate.
• VAR is a systems regression model (i.e. there is more than one
dependent variable) that can be considered a kind of hybrid between
the univariate time series models.

2
Few Assumption on VAR Model

• Treat all variables as endogenous. Helpful in a way that when you do


not want to explicitly specify “dependent” variable.
• Stationarity of the variables are important. If all variables are I(1) and
do not cointegrated, VAR at 1st difference is preferable. However, if all
variables are I(0), proceed with VAR at level.

3
Steps to VAR Modelling

• Step 1: Test order of integration (Unit root test)


• Test variables whether they are I(0) or I(1).
• Step 2: Setting the optimum lag length
• Important so that error terms do not suffer from non-normality,
autocorrelation, heteroscedasticity)
• Estimate VAR model with variables in levels. Choose lag that have
many asterisk(*) chosen by several criteria.

4
Steps to VAR Modelling

• Step 3: Estimate VAR Model


• Estimate VAR model using lag length that is chosen.
• Step 4: Granger Causality
• Restrict the variables to determine short-run granger causality

Notes: If all variables are I(1), check the cointegration between


variables. If they are cointegrated proceed with VECM otherwise VAR
at 1st difference.

5
VAR at Level Model

• The simple example of VAR model is bivariate VAR model.

𝑝 𝑝
𝑦𝑡 = 𝛽0 + σ𝑖=1 𝛽1𝑖 𝑦𝑡−𝑖 + σ𝑖=1 𝛽2𝑖 𝑥𝑡−𝑖 + 𝜀𝑡

𝑝 𝑝
𝑥𝑡 = 𝛽0 + σ𝑖=1 𝛼1𝑖 𝑥𝑡−𝑖 + σ𝑖=1 𝛼2𝑖 𝑦𝑡−𝑖 + 𝜀𝑡

6
VAR at 1st Difference Model

• The simple example of VAR model is bivariate VAR model.

𝑝 𝑝
∆𝑦𝑡 = 𝛽0 + σ𝑖=1 𝛽1𝑖 ∆ 𝑦𝑡−𝑖 + σ𝑖=1 𝛽2𝑖 ∆𝑥𝑡−𝑖 + 𝜀𝑡

𝑝 𝑝
∆𝑥𝑡 = 𝛼0 + σ𝑖=1 𝛼1𝑖 ∆𝑥𝑡−𝑖 + σ𝑖=1 𝛼2𝑖 ∆𝑦𝑡−𝑖 + 𝜀𝑡

7
Example: Bivariate VAR(1)
• A VAR model applies when each variable in the system does not only depend on its own
lags, but also the lags of other variables.
• A simple VAR example is:

8
VAR: Short-run Granger Causality

9
Short-run Granger Causality
• If past value of a variable or group of variables (say, 𝑥𝑡−𝑖 ) is found
to be helpful for predicting current value of another variable, or
group of variables (say, 𝑦𝑡 ), then 𝑥 is said to Granger-cause 𝑦.
• Granger causality distinguishes between unidirectional and bi-
directional causality.
• Unidirectional causality exists when 𝑥 causes 𝑦 but 𝑦 does not
cause 𝑥.
• If each of the variables causes the other, then a mutual feedback
exists between the variables.
• If there is no causality the series are said to be neutral.
10
Short-run Granger Causality

𝐻0 : 𝛽2𝑖 = 0 (no short-run causality from x to y)


𝐻1 : 𝛽2𝑖 ≠ 0 (there is short-run causality from x to y)
𝑖 = 1,2, …, p

𝐻0 : 𝛼2𝑖 = 0 (no short-run causality from y to x)


𝐻1 : 𝛼2𝑖 ≠ 0 (there is short-run causality from y to x)
𝑖 = 1,2, …, p
11
Q&A

12

You might also like