56
S. Prukumpai, et al
(IMW), real GDP (GDPR), real government spending
(GTR), short-term interest rate (INTS), long-term
government bond yield (INTL), and stock market
indices (SET). We used the real GDP and the SET
index to represent the real output and stock market,
respectively. In Thailand, increases in government
spending rather than tax cutting are typically used as a
scal policy mechanism. The short-term interest rate is
represented by the 1-day repurchase rate, which is used
as an instrument of monetary policy in Thailand. The
long-term interest rate (10-year government bond yield)
was included in the model to represent the transmission
channel of monetary policy and the crowding out effect
of scal policy. Finally, the world import value was
included to represent the external factor because the
international trade channel is important for ASEAN
economies. All data were collected from the CEIC
database at a quarterly frequency ranging from 1996
to 2017.
Econometric Methodology
Earlier, we discussed the complexity of the
interaction between the financial market, the real
economy, and economic policy. Therefore, the VAR
model is commonly employed to investigate the
dynamic relationships among real output, the stock
market, and monetary and scal policies. In the VAR
framework, the identication of shocks is crucial in
estimating the pattern of response of key variables to
shocks. Typically, the recursive method proposed by
Sims (1980) and the generalized method of Pesaran
and Shin (1998) are applied. However, in the case of
scal policy, a shock is dened as the changes in
government expenditures (or taxes) that are not due to
the business cycle.
While no consensus on the impact of scal policy
on economic activity has been concluded, researchers
generally agree on the linkage between scal and
economic activity. Besides the scal policy mechanism,
business cycle shocks also impact economic activity.
To handle these challenges, two main approaches are
applied: the narrative approach developed by Ramey
and Shapiro (1998) and the SVAR approach introduced
by Blanchard and Perotti (2002). The former assumes
that government spending is exogenous and orthogonal
to other information available at that time (Ramey &
Shapiro, 1998). The latter characterizes the dynamic
effects of shock in scal policy on economic activity
by using institutional features, that is, scal policy does
not respond to shocks that occur within the quarter
when using the quarterly data to achieve identication
(Blanchard & Perotti, 2002).
In this study, we followed the SVAR model. In
addition, the standard VAR models (reduced-form VAR)
with generalized impulse responses were also estimated
to check the robustness of the results. The details on
the econometric methodology are outlined as follows.
The structural VAR model. In this section, we
applied the structural VAR (SVAR) using the restrictions
suggested by Blanchard and Perotti (2002) and
Chatziantoniou et al. (2013). The representation of the
SVAR model of order has the following general form:
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
(1)
where
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
is a 6×1 vector of the endogenous variables,
that is,
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
= [IMW, GDPR, GTR, INTS, INTL, SET],
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
represents the 6×1 contemporaneous matrix,
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
is the 6×6 autoregressive coefcient matrix,
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
is
the 6×1 vector of structural disturbance, assumed
to have zero covariance. The covariance matrix of
the structural disturbances takes the following form:
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
. I n
order to estimate the SVAR model, the reduced form
was determined by multiplying both sides with
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
,
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
(2)
where
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
.
The reduced form has the covariance matrix of the
form
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
.
The structural disturbances can be derived by
imposing suitable restrictions on 0. In this study, the
short-run restrictions were set up as follows:
The structural VAR model. In this section, we applied the structural VAR (SVAR)
using the restrictions suggested by Blanchard and Perotti (2002) and Chatziantoniou et al.
(2013). The representation of the SVAR model of order has the following general form:
(1)
where
is a 6x1 vector of the endogenous variables, that is,
= [IMW, GDPR, GTR, INTS,
INTL, SET],
represents the 6x1 contemporaneous matrix,
is the 6x6 autoregressive
coefficient matrix,
is the 6x1 vector of structural disturbance, assumed to have zero
covariance. The covariance matrix of the structural disturbances takes the following form:
′
. In order toestimate the SVAR model, the
reduced form wasdetermined by multiplying both sides with
(2)
where
and
.The reduced form has the covariance matrix
of the form
′
The structural disturbances can be derived by imposing suitable restrictions on
In
this study, the short-run restrictions wereset up as follows:
(3)
The restrictions in the SVAR model wereimposed based on severalprinciples. First,
income contemporaneously reacts to external shocks but is not concurrently influenced by
other factors in the model. However, the GDP is the important factor that affectsthe long-term
(3)
The restrictions in the SVAR model were
imposed based on several principles. First, income
contemporaneously reacts to external shocks but is