HUL_715_Presentation_final.pdf

Introduction
Methodology
Conclusion
Export-led growth in Bangladesh: a time
series analysis
K. A. Al Mamun1
H. K. Nath2
1Department of Economics,
Southern Methodist University, Dallas, USA
2Department of Economics and International Business,
Sam Houston State University, Huntsville, USA
November 14, 2022
Vinay Tomar (2021HES7092) Export-led growth in Bangladesh:a time series analysis 1/15

Introduction
Methodology
Conclusion
Introduction
Data
Introduction
In recent years, Bangladesh has experienced not only a
substantial increase in the volume of exports but also
important changes in the composition of those exports.

Introduction
Methodology
Conclusion
Introduction
Data
Introduction
It is moving away from traditional items such as jute and
jute products and towards newly manufactured products
such as ready-made garments.

Introduction
Methodology
Conclusion
Introduction
Data
Introduction
It is moving away from traditional items such as jute and
jute products and towards newly manufactured products
such as ready-made garments.
So, this paper presents time series analysis of the
export–output relationship for Bangladesh.

Introduction
Methodology
Conclusion
Introduction
Data
Data
Data set is taken from the International Financial
Statistics published by the IMF comprised of:

Introduction
Methodology
Conclusion
Introduction
Data
Data
Quarterly data on industrial production index

Introduction
Methodology
Conclusion
Introduction
Data
Data
Exports of goods and services

Introduction
Methodology
Conclusion
Introduction
Data
Data
Exports of goods only for a period from 1976:1 to 2003:3

Introduction
Methodology
Conclusion
Introduction
Data
Data
Exports of goods only for a period from 1976:1 to 2003:3
For all three data series, the base year is 2000 and the
export values are given in constant US dollars

Introduction
Methodology
Conclusion
Unit root test
Engle-Granger cointegration test
VECM and Granger causality
Augmented Dickey-Fuller Test
ADF tests were carried out to find out the order of
integration for each of the above three series.

Introduction
Methodology
Conclusion
Unit root test
First, tests in levels and then in first differences were
carried out. Each series started with the most flexible
specification of the test equation that includes an
intercept and a trend :
∆zi,t = αi,o + αi,1t + γzi,t−1 +
p
X
j=1
βi,j ∆zi,t−j + εi,t

Introduction
Methodology
Conclusion
Unit root test
First, tests in levels and then in first differences were
carried out. Each series started with the most flexible
specification of the test equation that includes an
intercept and a trend :
∆zi,t = αi,o + αi,1t + γzi,t−1 +
p
X
j=1
βi,j ∆zi,t−j + εi,t
The ADF test is essentially the test of significance of the
coefficient γ in the above equation.

Introduction
Methodology
Conclusion
Unit root test
In order to select the lag length p, we start with a
maximum lag of 8 and pare it down to the appropriate lag
by examining the AIC.

Introduction
Methodology
Conclusion
Unit root test
If we do not find the intercept and the trend – both or
one of them – to be statistically significant at the 10 per
cent significance level, we drop the insignificant term(s)
and re-estimate the test statistics.

Introduction
Methodology
Conclusion
Unit root test
If we do not find the intercept and the trend – both or
one of them – to be statistically significant at the 10 per
cent significance level, we drop the insignificant term(s)
and re-estimate the test statistics.
The results are reported in Table 1. The number of lags
of the augmented terms and other specifications of the
test equation is included in the table. We’ll see, all three
series are integrated of order one or I(1)

Introduction
Methodology
Conclusion
Unit root test
Result
(Y) (XG + XS ) (XG )
Panel A: In Levels
Intercept term? Yes Yes Yes
Time trend? Yes Yes Yes
Lag length 4 4 5
ADF test statistics 1.98 1.94 2.05
MacKinnon’s p-value 0.60 0.62 0.57
Panel B: In Differences
Intercept term? Yes Yes Yes
Time trend? No No No
Lag length 3 3 2
ADF test statistics 5.93 5.98 10.65
MacKinnon’s p-value 0.00 0.00 0.00

Introduction
Methodology
Conclusion
Unit root test
Engle-Granger cointergration test
Intuition: If variables are cointegrated, then the residual
of the cointegrating regression should be stationary.

Introduction
Methodology
Conclusion
Unit root test
So, we test the residuals for stationarity obtained from an
OLS regression of industrial production on
contemporaneous values of exports:
yt = β0 + β1xt + et
ˆ
et = yt − ˆ
β0 − ˆ
β1xt

Introduction
Methodology
Conclusion
Unit root test
So, we test the residuals for stationarity obtained from an
OLS regression of industrial production on
contemporaneous values of exports:
yt = β0 + β1xt + et
ˆ
et = yt − ˆ
β0 − ˆ
β1xt
Above equation represents a long-run equilibrium
relationship between industrial production and exports,
here yt and xt are non-stationary series.

Introduction
Methodology
Conclusion
Unit root test
Results
Panel A: (XG + XS ) Estimated long-run
(yt = 1.34 + 0.43xt + et)
Dickey-Fuller Test on {êt} Lag = 0 Lag = 5
Test statistic −5.96 −1.93
5% critical value −3.37 −3.17
Panel B: (XG ) Estimated long-run
(yt = 1.48 + 0.42xt + et)
Dickey-Fuller test on {êt} Lag = 0 Lag = 5
Test statistic −6.67 −1.99
5% critical value −3.37 −3.17

Introduction
Methodology
Conclusion
Unit root test
These tests, however, give contradictory results:

Introduction
Methodology
Conclusion
Unit root test
The test with no lag indicates that the residuals are
stationary and the two variables are cointegrated.

Introduction
Methodology
Conclusion
Unit root test
The test with no lag indicates that the residuals are
stationary and the two variables are cointegrated.
The test with 5 lags (selected using AIC) indicates that
they are not stationary, and two variables are not
cointegrated.

Introduction
Methodology
Conclusion
Unit root test
Vector error correction model (VECM) and GC
The existence of a cointegrating relationship suggests
that we should estimate a VECM of industrial production
and exports, as represented by the equation:
∆zi,t = δi,o + δi,1êt−1 +
2
X
i=1
p
X
j=1
λi,j ∆zi,t−j + ξi,t
∆Yt = ΦDt +ΠYt−1 +Γ1∆Yt−1 +· · ·+Γp−1∆Yt−p+1 +ϵt

Introduction
Methodology
Conclusion
Unit root test
Vector error correction model (VECM) and GC
The existence of a cointegrating relationship suggests
that we should estimate a VECM of industrial production
and exports, as represented by the equation:
∆zi,t = δi,o + δi,1êt−1 +
2
X
i=1
p
X
j=1
λi,j ∆zi,t−j + ξi,t
∆Yt = ΦDt +ΠYt−1 +Γ1∆Yt−1 +· · ·+Γp−1∆Yt−p+1 +ϵt
Π captures adjustments towards the long-run equilibrium
and contains the cointegrating relationships, Γk captures
short-run deviations from the equilibrium and Dt contains
deterministic terms.

Introduction
Methodology
Conclusion
Unit root test
The coefficient δi,1 represents the long-run causal effect in
relation to the long-run equilibrium relationship of the
cointegrated processes.

Introduction
Methodology
Conclusion
Unit root test
The coefficients of the lagged values:

Introduction
Methodology
Conclusion
Unit root test
λ2,j s, represent short-run effects of exports on industrial
production.

Introduction
Methodology
Conclusion
Unit root test
production.
λ1,j s, represent short-run effects of industrial production
on exports.

Introduction
Methodology
Conclusion
Unit root test
production.
λ1,j s, represent short-run effects of industrial production
on exports.
A test of joint significance of these lagged terms
constitutes a short-run Granger causality test. The results
are reported in Table 3.

Introduction
Methodology
Conclusion
Unit root test
(∆y) (∆x)
Panel A: (XG + XS )
Estimated value of δi,1 −0.12 0.09
t-statistics −1.88∗
0.85
H0 : δi,1 = 0 3.53∗
0.73
H1 : δi,1 ̸= 0
H0 : λi,j = 0 for all j 0.75 0.80
H1 : λi,j ̸= 0 for at least one j
Panel B: (XG )
Estimated value of δi,1 −0.15 0.10
t-statistics −2.09 ∗ ∗ 0.69
H0 : δi,1 = 0 4.36∗∗
0.47
H1 : δi,1 ̸= 0
H0 : λi,j = 0 for all j 0.95 0.96
H1 : λi,j ̸= 0 for at least one j

Introduction
Methodology
Conclusion
Unit root test
Results
The estimated value of the speed of adjustment
coefficient is negative, −0.12 and statistically significant,
−1.88∗
in the industrial production equation (∆y). It
indicates that the further away industrial production
deviates from its long-run relationship with exports, the
lower the growth rate and vice versa.

Introduction
Methodology
Conclusion
Unit root test
Results
−1.88∗
The speed of adjustment coefficient, δi,1 in the the export
equation (∆x) is statistically insignificant in both of the
two cases: F stat- 0.73 and 0.47!

Introduction
Methodology
Conclusion
Unit root test
Results
−1.88∗
The speed of adjustment coefficient, δi,1 in the the export
equation (∆x) is statistically insignificant in both of the
two cases: F stat- 0.73 and 0.47!
The short-run Granger causality test results reported in
the last row of each panel in Table 3 (H1) indicates that
there is no causal relationship between export growth and
industrial growth.

Introduction
Methodology
Conclusion
Conclusion
Conclusion
While the analysis suggests that:

Introduction
Methodology
Conclusion
Conclusion
Conclusion
There is a positive long-run equilibrium relationship
between exports and industrial production, there is no
evidence of a short-run causal relationship between these
two variables.

Introduction
Methodology
Conclusion
Conclusion
Conclusion
There is a positive long-run equilibrium relationship
between exports and industrial production, there is no
evidence of a short-run causal relationship between these
two variables.
Furthermore, the long-run causality seems to run from
exports to industrial production.

Introduction
Methodology
Conclusion
Conclusion
Thank you!

HUL_715_Presentation_final.pdf

Recommended

Recommended

More Related Content

Similar to HUL_715_Presentation_final.pdf

Similar to HUL_715_Presentation_final.pdf (20)

Recently uploaded

Recently uploaded (17)

HUL_715_Presentation_final.pdf