Scribd
Upload
62 views4 pages

Ezekiel Thesis VAR Model Interpreted

The document describes a vector autoregressive (VAR) model that includes 2 lags each of GDP growth, unemployment rate, and fertility rate. It analyzes the model results including the impulse response functions and forecast variance decomposition. The impulse response function shows GDP growth decreases after 1 year due to a fertility rate shock, while unemployment is unaffected. GDP growth and unemployment rate mostly explain their own future variances, while fertility rate explains around 30% of GDP growth variance.

Uploaded by

Ezekiel Fernandez
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
62 views4 pages

Ezekiel Thesis VAR Model Interpreted

The document describes a vector autoregressive (VAR) model that includes 2 lags each of GDP growth, unemployment rate, and fertility rate. It analyzes the model results including the impulse response functions and forecast variance decomposition. The impulse response function shows GDP growth decreases after 1 year due to a fertility rate shock, while unemployment is unaffected. GDP growth and unemployment rate mostly explain their own future variances, while fertility rate explains around 30% of GDP growth variance.

Uploaded by

Ezekiel Fernandez
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

VECTOR AUTOREGRESSIVE MODEL

The final VAR model is shown in the following table.


Table __: Vector Autoregressive Model

GDP Growth Unemployment


Rate Rate (Diff) Fertility Rate

GDP Growth Rate (lag 1) -0.257738 -0.291143 -0.001053


(0.28176) (0.15626) (0.00057)
[-0.91476] [-1.86325] [-1.84823]

GDP Growth Rate (lag 2) -0.115026 -0.192435 -0.000909


(0.26280) (0.14574) (0.00053)
[-0.43769] [-1.32036] [-1.71086]

Unemployment Rate (Diff)


(lag 1) -0.431284 -0.479736 0.000549
(0.48798) (0.27062) (0.00099)
[-0.88381] [-1.77271] [ 0.55615]

Unemployment Rate (Diff)


(lag 2) -0.483181 -0.130191 0.000317
(0.48406) (0.26845) (0.00098)
[-0.99818] [-0.48497] [ 0.32388]

Fertility Rate (lag 1) 23.61609 25.44061 1.795534


(34.7259) (19.2582) (0.07019)
[ 0.68007] [ 1.32103] [ 25.5802]

Fertility Rate (lag 2) -26.36259 -26.03084 -0.817338


(33.8280) (18.7602) (0.06838)
[-0.77931] [-1.38755] [-11.9533]

C 17.58854 5.381355 0.073248


(7.79998) (4.32568) (0.01577)
[ 2.25495] [ 1.24405] [ 4.64585]

R-squared 0.329698 0.449880 0.999914


Adj. R-squared -0.035921 0.149814 0.999867
Sum sq. resids 34.04966 10.47215 0.000139
S.E. equation 1.759382 0.975712 0.003556
F-statistic 0.901753 1.499272 21359.25
Log likelihood -31.27793 -20.66602 80.39413
Akaike AIC 4.253103 3.074002 -8.154903
Schwarz SC 4.599359 3.420258 -7.808647
Mean dependent 5.298301 -0.227778 3.300778
S.D. dependent 1.728609 1.058192 0.308787

Determinant resid covariance (dof adj.) 2.32E-05


Determinant resid covariance 5.29E-06
Log likelihood 32.72969
Akaike information criterion -1.303299
Schwarz criterion -0.264532
Number of coefficients 21
The final model includes 2 lags of each endogenous variables—VAR(2). To interpret the results of the
VAR model, we will look into two functions that defines the forecasting ability of the model—the
Impulse Response Function (IRF) and the Forecast Variance Decomposition.

Impulse response function gives the effect of a one-time shock in fertility rate to GDP growth rate and the
change in unemployment rate. We say that the effect is significant if the horizontal line for 0 is not within
the bounds of the standard errors (red dotted lines). The Cholesky ordering considered was fertility rate >
unemployment rate > GDP growth rate.

The figure below shows the IRF of GDP growth rate to fertility rate along with the corresponding 2
standard error boundary up to 12 lags.

Figure __: Impulse Response Function of GDP Growth Rate to Fertility Rate

A one-time shock in fertility rate has significant effect on GDP growth rate on the first lag. Specifically,
GDP growth rate will decrease after a year if the fertility rate of the country will suddenly increase this
year. Then, this one-time shock will have no effect on the succeeding years later on. Let’s go now to the
IRF of unemployment rate (differenced) to fertility rate.

Figure __: Impulse Response Function of Unemployment Rate (diff) to Fertility Rate

Based on the figure above, a one-time shock in fertility rate has no significant effect in the change of
unemployment rate. This result further supports the earlier OLS model for differenced unemployment rate
and fertility rate.

Next thing, we will look into the forecast variance decomposition. Variance decomposition gives the
percentage of the total variance explained by the independent variable on the future movement of the
dependent (outcome) variable.

Table __: Variance Decomposition of GDP Growth Rate

GDP Growth Unemployment Fertility


Period Rate Rate (diff) Rate

1 68.63059 0.000000 31.36941


2 64.97986 4.719858 30.30028
3 64.56884 5.246873 30.18429
4 63.69962 6.528648 29.77173
5 63.23986 7.223421 29.53672
6 63.21392 7.233724 29.55235
7 63.16316 7.253863 29.58298
8 63.11853 7.307684 29.57378
9 63.08180 7.300870 29.61733
10 63.04389 7.293577 29.66253
11 63.00226 7.301969 29.69578
12 62.96099 7.296524 29.74249

As shown in the table above, majority of the future variance of GDP growth rate are explained by itself.
More than 60% of the forecast error variance is explained by GDP growth rate itself. Meanwhile, fertility
rate has a significant contribution in explaining the forecast error variance of GDP growth rate. It
averages by about 30% across 12 years.

Table __: Variance Decomposition of Differenced Unemployment Rate

GDP Growth Unemployment


Period Rate Rate (diff) Fertility Rate

1 4.233022 90.56796 5.199020


2 11.11035 79.67016 9.219488
3 10.87705 80.01840 9.104545
4 10.86474 80.08213 9.053136
5 11.55186 78.90011 9.548023
6 11.52934 78.93856 9.532102
7 11.52627 78.92719 9.546541
8 11.57334 78.84457 9.582085
9 11.57032 78.84983 9.579854
10 11.57095 78.84859 9.580462
11 11.57349 78.84596 9.580554
12 11.57352 78.84449 9.581989

As shown in the table above, majority of the future variance of change in unemployment rate are
explained by itself. Up to 90% of the forecast error variance is explained by differenced unemployment
rate itself in the first lag then it decreased to an average of 79% in the succeeding years. Meanwhile,
fertility rate has no significant contribution in explaining the future forecast error variance for
unemployment rate (differenced).

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy