Bank, Stock Market And Economic Growth: Bangladesh

Perspective
Prashanta Kumar Banerjee, Md. Nehal Ahmed, Md. Mosharref Hossain

The Journal of Developing Areas, Volume 51, Number 2, Spring 2017, pp.
17-29 (Article)

Published by Tennessee State University College of Business

DOI: https://doi.org/10.1353/jda.2017.0028

For additional information about this article

https://muse.jhu.edu/article/657925

Access provided by Australian National University (11 Jul 2018 22:38 GMT)
 The Journal of Developing Areas
Volume 51 No. 2 Spring 2017

BANK, STOCK MARKET AND ECONOMIC

GROWTH: BANGLADESH PERSPECTIVE
Prashanta Kumar Banerjee
Md. Nehal Ahmed
Md. Mosharref Hossain
Bangladesh Institute of Bank Management (BIBM), Bangladesh

ABSTRACT

The paper empirically investigates the relationship between bank, stock market and economic
growth in Bangladesh. In investigating empirical relationship , GDP at the current price,
private sector credit given by banks and market capitalization are considered as indicators of
economic growth, banks’ development and stock market development, respectively. Three
possible regression models are estimated to know the relationship between economic growth
and financial sector development. The paper shows that bank’s credit to the private sector is
both positively and robustly contributing to the economic development of the country when
bank credit to private sector entered alone in regression as an independent variable. Besides
long-term relationship, contemporaneous change in bank’s credit to the private sector has
profound positive short-run feedback effects to the economic growth. In assessing impact of
stock market development on economic growth, no significant long-run relationship is found
between stock market development and economic growth in Bangladesh. However, net
positive subdued short-term effect of stock market development is evident. But when both
bank and equity market jointly enter the model to find out the relationship of both variables
with economic growth, a long-run relationship has been evident without statistical
significance. It indicates that Bangladeshi financial system comprising both banking sector and
stock market jointly is not still a strong promoter of its economic growth, although banking
development alone is robustly associated with economic development.Bangladesh therefore
needs to enhance the efficiency of banks for increasing credit to the private sector. An
immediate initiative is also required to make both equity and debt market as regular sources of
finance for the economy. In this perspective, ensuring smooth operation of primary and
secondary market, increasing financial literacy among investors, minimizing volatility of the
market, expanding issuer base, creating both individual and institutional investors, enhancing
efficiency of the brokerage house, adding innovative financial services, initiating knowledge
based trading, introducing shelf registration system, creating more professional trustee and
ensuring authenticate credit rating from the rating agencies are required to be ensured.

JEL Classifications: E44,C22, G24,G21

Keywords: Bank Credit to Private Sector, Equity Market, Economic Growth, Cointegration
Corresponding Author's Email Address: banerji611@gmail.com

INTRODUCTION

A large body of literature indicates positive influence of the development of a

country’s financial sector on its economic growth. Empirical findings seem
consistent with this argument (Schumpeter 1912, Levine and Sarah Zervos 1998,
Arestis et al. 2001). Levine and Zervos (1998), Rajan and Zingles (1998) and Beck
and Levine (2004) demonstrate the importance of both banks and equity markets for
economic growth. Herring and Chatusripitak (2001) note that the absence of a bond
market may render an economy less efficient and significantly more vulnerable to
financial crisis.
 18

Is there any predicament if financial system of a country is heavily

dependent on a single source of finance particularly on bank based finance? The
financial turmoil that occurred in East Asia in mid-1997 taught policy makers,
academics and practitioners that excessive reliance on banks as the vehicle through
which savings are channeled to investment projects significantly exacerbates
economic downturn when the banking sector suffers a crisis. Park and Oh (2006)
point out that one of the root causes of the financial crisis in 1997-1998 was heavy
dependent on banking systems to finance domestic investment. The latest financial
crisis in Europe also puts finger on more dependency of economy on bank finance.
Europe’s economy is more dependent on banks than the capital markets. European
banks account for around 75 per cent of corporate financing, compared with about 30
per cent in the US (European Banking Federation 2012). On the other hand, the
experience of the United States securities market during financial crises – one
resulting from the Latin American debt crisis in the 1980s, the other from the real
estate crisis of 1990 and the latest sub-prime crisis that began in December 2007
ending in June 2009 – is a constructive and shining example and it showed how
securities markets comprising of equity and debt securities can provide the corporate
sector with alternative sources of financing. During these periods, the US banking
sector suffered large losses that reduced its capital base drastically and severely
curtailed its ability to lend. The ensuing liquidity crunch substantially reduced the
bank credit to US corporations. The US domestic securities markets, to varying
degrees, functioned as alternative sources of financing for corporations when the
banking sector was under pressure. Greenspan (2000) stresses the importance of
having multiple avenues of financial intermediation, which served the US well during
the credit crunch of the late 1980s when debt markets substituted for the loss of bank
financial intermediation in a banking crisis related to the real estate cycle.
Bangladesh economy distinctly depends on bank dominated financial
system. Direct financing through issuing shares is insignificant and the other direct
financing mechanism through issuing corporate bonds here is almost nonexistent. In
the bank based financing system of all economies, maturity mismatch between assets
and liabilities is an inherent problem. Additionally, lack of expected efficiency of
intermediation is also a cause of concern in Bangladesh. Islam (2012) asserts that
there are indications of collusive, oligopolistic fixing of lending rates through the
association of bank officials and chief executives. These limitations of the banking
sector of Bangladesh, which may be common features in other developing countries,
increase the importance of having a sound and organized capital market comprising
of both equity and debt market for fulfilling the needs of financing business
activities.
The capital market encourages specialization as well as acquisition and
dissemination of information, thereby reducing the cost of mobilizing savings and
facilitating investment (Greenwood and Smith, 1997). But the Bangladesh equity
market, to say the least, till today is not broad or deep enough. Of around 10,000
public limited companies of the country, only 10 per cent are listed with the
exchange (The Financial Express, July 04, 2013). In this perspective, knowing the
relationship between the conduits of finance of the country with economic growth is
important to make financial sector more diversified and efficient for facilitating
sustainable and balanced economic growth of the country. However, financing
through corporate bond market is not considered here as it is almost absent here. The
main objective of this endeavor is to investigate the empirical relationship between
bank, stock market and economic growth in Bangladesh. The remainder of the paper
proceeds as follows: Section II briefly reviews the related literature. Section III states
 19

methodology, data sources and outlines framework for empirical analysis. Section IV
shows results and interpretation of empirical evidences. Section V offers conclusions
and remarks.

LITERATURE REVIEW

Economic development of a country depends on many factors. Both bank and stock
market provides important contribution to the growth across the globe. Several
studies have been conducted to find the role of bank and stock market on economic
growth. The study conducted by Levine and Zervos in 1998 is considered as seminal
one in this area. They empirically assess the impact of stock markets and banks on
long-run economic growth using an endogenous growth model. After examining data
on 47 countries over a period of 1976 to 1993, the results show that both stock
markets and banking development are positively and significantly related to
economic growth and both are good predictors of economic growth.
There are at least four ways in which financial system development
contributes to economic growth. They are extensively described in the surveys
provided by Pagano (1993) and Levine (1997, 2002). First, financial intermediaries
may lower the costs of gathering and processing information and thereby improve the
allocation of resources (Boyd and Prescott, 1986). Such improvement of information
about all economic agents can boost economic growth. Second, Bencivenga and
Smith (1993) show that banks that alleviate the corporate governance problem by
lowering monitoring costs will reduce credit rationing and thereby spur growth.
Third, financial intermediaries and security markets provide vehicles for trading,
pooling and diversifying risk. Fourth, financial systems that encourage the
mobilization of savings by providing attractive instruments and saving vehicles can
significantly affect economic development.
Hondroyiannis et. al. (2005) show that both banks and stock markets
financing can promote economic growth in the long-run although their effect is
small. Furthermore, the contribution of stock markets in financing economic growth
appears to be substantially smaller compared to bank finance. Asante et. al. (2011
find that bank competition and stock market development cause economic growth in
Ghana. Also, in the long run, banking competition is good for economic growth.
However, there is a disproportionate response of economic growth to stock market
development. Atje and Jovanovic (1993) conclude that stock markets have long-run
impacts on economic growth and it is also found that stock markets manipulate
economic growth through a number of channels that are liquidity, risk
diversifications, acquisition of information about firms, corporate governance and
savings mobilization (Levine & Zervos 1996).
Antonios (2010) investigates the causal relationship between stock market
development and economic growth for Germany for the period 1965-2007 using a
Vector Error Correction Model (VECM). The results of the tests indicated that there
is a unidirectional causality between stock market development and economic growth
with direction from stock market development to economic growth. Above literature
supports that both bank and stock market separately plays a significant role in the
economic development of a country. But knowing joint impact of both bank and
stock market on the economic development is equally important. There are studies
which provide the evidence of both bank and stock market involvement in the overall
development process. Levine and Zervos (1993) investigate whether banking and
stock market indicators are both robustly correlated with current and future rates of
economic growth, capital accumulation, productivity growth and private saving.
 20

They found that stock market liquidity is positively and significantly correlated with
current and future rates of economic growth, capital accumulation, and productivity
growth. Moreover, the level of banking development as measured by bank loans to
private enterprises divided by GDP also showed the positive relationship
significantly. They conclude that banking development and stock market liquidity are
both good predictors of economic growth, capital accumulation, and productivity
growth.
Rousseau and Wachtel (2000) use panel vector auto-regression with
generalized method of moment technique to examine simultaneously the relationship
between stock markets, banks and economic growth. After examining the
relationship on 47 countries using annual data from 1980-1995, their results indicate
that both banks and stock markets promote economic growth. Arestis et al. (2001)
use quarterly data on five developed countries and find that both banks and stock
markets development lead to economic growth. They also find that the impact of
banking sector development is substantially larger than that of stock market
development. Boyd and Smith (1998) and Blackburn et. al. (2005) have all shown
that both stock market and banks are necessary in promoting economic growth. They
consider stock markets as compliment to banks rather than substitutes. Beck and
Levine (2004) assess whether stock markets and banks have positive influence on
economic growth. Using a dynamic panel data set on 40 countries over a period
1976-1998 and with the application of GMM estimators their results shows that both
stock market and financial development enter all of the system panel growth
regression significantly.

METHODOLOGY, DATA SOURCES AND FRAMEWORK FOR

EMPIRICAL ANALYSIS

Methodology and Data Sources

In investigating empirical relationship between economic growth and both bank and
stock market development, we first consider GDP at current prices as an indicator of
economic growth. Then we consider stock data of private sector credit given by
banks as an indicator of banks’ development and stock data of market capitalization
as an indicator of stock market development. Arestis et al. (2001) note that stock data
are likely to have more time series property that makes it suitable for cointegration
analysis. GDP is considered at current prices as all other data are also available at
current prices only. Credit to private sector by banks is considered as an indicator of
banks’ development by following Levine and Zervos (1998). Market capitalization is
contemplated here to measure equity market contribution considering the theory that
a larger share market contributes more to a growing and large economy. Levine and
Zervos (1998) show that market capitalization is not a good predictor of economic
growth. However, Beck and Levine (2004), and Tang (2006) subsequently use
market capitalization in their studies. In addition, Arestis et al. (2001) highly
recommend market capitalization for time series analysis.
Annual stock data from 1974 through 2012 have been used when
econometric relationship is examined between economic growth and banks’ credit to
the private sector as data of both variables are available since 1974. However, data on
market capitalization of Stock Exchanges is available only since 1988. For
determining empirical relationship between economic growth and stock market
development, if we take data 1988 onwards number of observations is supposed to be
very small. Any econometric model based on time series data suffers from degrees of
 21

freedom problem if a few observations have been taken into consideration. To find
out good results through using sufficient observations, quarterly data of market
capitalization have been considered. However, in absence of availability of quarterly
data of GDP in Bangladesh, GDP data has been converted into quarterly data by
following Linear-match last method with a view to matching with the quarterly data
of market capitalization. Then econometric model between economic growth and
stock market development has been run on the basis of the converted data. Finally, in
case of finding relationship between economic growth and both bank and stock
market, quarterly data have been considered for all three variables. As quarterly data
for bank’s credit to the private sector and market capitalization are available, we use
only converted quarterly data for GDP for the estimation. Data are gleaned from the
various issues of Annual Report of Bangladesh Bank and Economic Trends and
Scheduled Banks Statistics published by the Bangladesh Bank.

Empirical Design

Three possible regression models are estimated here as the basis to know the
relationship between economic growth and financial sector development. The first
equation is estimated to find out the relationship between GDP at current prices and
banks’ private sector credit to know the empirical relationship between economic
growth and banking sector development. With a view to estimating the relationship
between economic growth and stock market development, the second equation is on
GDP at current prices and market capitalization of stock exchanges in Bangladesh.
Finally, relationship has been examined in one equation by taking both bank private
sector credit and stock market development as independent variables and economic
growth as dependent variable with a view to understating the collective contribution
of both variables to economic growth of Bangladesh. The estimating regression
models are specified as follows:

LnYt=  + 1LnBPSCt + et (1)

LnYt=  + 1Ln CAPIt + et (2)

LnYt=  + 2LnBPSCt +3Ln CAPIt + et (3)

Where Y = GDP of Bangladesh at current prices, BPSC = Banks’ private sector credit
and CAPI = Market capitalization in stock exchanges. All variables are converted
into natural logs. The expected signs of the parameters are: > 0, 1> 0, 1> 0, 2> 0
and 3> 0.
First, the time series property is investigated by implementing the ADF
(Augmented Dicky–Fuller) for the unit root (nonstationarity) following (Dickey and
Fuller, 1981; Fuller, 1996). The KPSS (Kwiatkowski, Philips, Schmidt and Shin) test
for no unit root (stationarity) is applied as a counterpart of ADF following
(Kwiatkowski, et al., 1992). Second, in the event of nonstationarity of each variable,
the co integrating relationship among variables is studied by the Johansen-Juselius
procedure (Johansen 1991; Johansen and Juselius 1992, 1990) to overcome the
associated problem of spurious correlation and misleading inferences. In this
procedure, all the variables must have the same order of integration. However, to
address the issue of unequal order of integration , the ARDL model or bounds-testing
procedure, as suggested by Pesaran et al. (2001) is needed to be applied.
 22

Δ𝑙𝑛𝑌𝑡 = α0 + ∑𝑝𝑖=1 b∆lnYt−i + ∑𝑝𝑖=0 c∆ ln BPSCt−i +

∑𝑝𝑖=0 d∆lni CAPIt−i+𝜆1 𝑙𝑛𝑌𝑡−1 +𝜆2 𝑙𝑛𝐵𝑃𝑆𝐶𝑡−1 +
𝜆3 𝑙𝑛𝐶𝐴𝑃𝐼𝑡−1 + 𝜀𝑡 (4)

The null and the accompanying alternative hypotheses for the cointegrating
relationship are:
Ho: 𝜆1 = 𝜆2 =𝜆3 = 0 for no cointegration
Ha: 𝜆1 ≠ 𝜆2 ≠ 𝜆3 ≠ 0 for cointegration

If the calculated F-statistic is above its upper critical value, the null
hypothesis of no long-run relationship can be rejected irrespective of the orders of
integration for the time series variables. Finally, on the evidence of cointegrating
relationship, a vector error-correction model (VECM) is estimated for long-run
causality and short-term dynamics2.

The VECM following Engle and Granger (1987) is specified as follows:

n m
ΔlnYt = α + ECM t 1 +  bi ln Y t i +  ci  ln BPSC t i + Ut
i 1 i 0
(5)

n m
ΔlnYt =  1 + ECM t 1 +  di ln Y t i +  ei  ln BPSC t i +
i 1 i 0
k
 fi ln CAPI t i + U1t (6)
i 0
A negative and statistically significant coefficient of ECM t–1 supports long-run
causality that springs from the independent variables. The short-term effects are
inferred by the sizes and signs of bis, cis , dis, eis, and fis as well as the statistical
significance of the overall calculated F-statistic. If there is no evidence of
cointegration, VAR is estimated.
n m
ΔlnYt =  ii +  gi ln Y t i +  hi ln CAPI t i + Uiit (7)
i 1 i 0

The Results and Discussion

First, for each variable, the results of the ADF and KPSS tests with orders of
integration are reported in Tables 1 and 2. All variables are found to be non-
stationary based on both the ADF and KPSS tests. The same order of integration was
found on the first differencing i.e. I(1) for LnY and LnBPSC when annual data for
both variables were considered (Table 1). In case of consideration of quarterly data,
LnY and LnCAPI show same order of integration on the second differencing i.e. I(2)
(Table 2). Same order of integration justifies the implementation of Johansen-
Juselius procedure. Accordingly, the study implements Johansen-Juselius procedure
for searching co integration among the variables for equations 1 and 2. But quarterly
data of LnBPSC is found to be stationarity in First differencing I(1) (Table 2). The
ambiguity of the order of integration for variables used in equation 3, where
relationship between bank, stock market and economic growth is investigated, lends
 23

support to the use of Auto Regressive Distributed Lag Model (ARDL) bounds
approach to cointegration.

TABLE 1. ADF AND KPSS TEST RESULTS AND ORDER OF

INTEGRATION:ANNUAL DATA

Variables ADF KPSS

Level 1st 2nd Level 1st
Difference Difference Difference
LnY 0.350881 -11.49792* - 0.730481* -
LnBPSC -2.832063 -3.390571* - 0.706510* -
Note: The Mackinnon (1996) critical values are –3.653730 and –2.957110 at 1%
and 5% levels of significance, respectively. The KPSS critical values (Kwiatkowski et
al., 1992, Table 1) are 0.73900 and 0.46300 at 1% and 5% levels of significance,
respectively. Asterisk (*) indicates stationarity of the variables.

TABLE 2. ADF AND KPSS TEST RESULTS AND ORDER OF

INTEGRATION: QUARTERLY DATA

Variables ADF KPSS

Level 1st 2nd Level 1st
Difference Difference Difference
LnYQ 2.300545 -0.893930 -4.986268* 1.215917 0.472580*
LnBPSCQ 0.346957 -3.583240* 1.189994 0.11029*
LnCAPIQ 0.256652 -2.236624 -7.287262* 1.003690 0.275252*
Note: The Mackinnon (1996) critical values are –3.653730 and –2.957110 at 1%
and 5% levels of significance, respectively. The KPSS critical values (Kwiatkowski et
al., 1992, Table 1) are 0.73900 and 0.46300 at 1% and 5% levels of significance,
respectively. Asterisk (*) indicates stationarity of the variables.

Second, The trace AND max test results calculated in Johansen and Juselius procedures
are reported in Table-3. As observed in the table there is an evidence of co
integration relationship between LnY and LnBPSC which are captured in equation 1
in terms of both trace AND max tests. As these variables are co integrated, this indicates
a long-run relationship exists between economic growth and bank finance. In this
case, the vector error correction model (VECM) - as given in equation (5) – is
estimated to capture the both short term and long term dynamics.
Although LnBPSC conintegrated with LnY (Equation 1) but no
cointegrating relationship has been found when LnY and LnCAPI have been
considered as per equation 2 and consequently, the Vector Autoregressive (VAR)
model has been estimated to capture the short term dynamics by following equation
7. For equation 3 where both bank and equity finance have been considered as
independent variables and economic growth as dependent variable, as per ARDL
model, calculated F- statistics is 33.12683 which is higher than the upper value of
4.01 at the 5 per cent level of significance (Table 4). Moreover, none of the estimated
coefficients of LnY, LnBPSC and LnCAPI as represented by 𝜆1 , 𝜆2 and 𝜆3
respectively is equal to 0.
 24

TABLE 3. COMPUTED VALUE OF TRACE AND MAX STATISTICSa

Hypotheses Equation 1 Equation 2 0.05 Critical

Value
LnY=  + LnBPSC + Ut
LnY=  + Ln CAPI
+ Ut
COMPUTED VALUE OF trace STATISTICS
None 48.22710b 11.28944 15.50
(H0: r = 0)
At most 1 2.189086 4.321702 3.84
(H0: r ≤ 1)
COMPUTED VALUE OF max STATISTICS
None 46.03801b 6.967740 14.27
(H0: r = 0)
At most 1 2.189086 4.321702 3.84
(H0: r ≤ 1)
a λtrace test indicates cointegrating equations at the 0.05 level and λmax test indicates co
integrating equations at the 0.05 level.
b Denotes rejection of the null hypothesis of no co integration at the 0.05 level.

This implies that the alternative hypothesis of co-integration may be accepted. It is

concluded that there prevails a co-integration relationship among the variables
considered in the equation 3. Given the above information about the presence of
cointegration, we can examine long run association and ECM version under ARDL
procedures. (see, Pesaran et. al. 2001; Pesaran and Shin 1999 for details).

TABLE 4. ARDL COINTEGRATION STATISTICS

Dependent F-Statistics Probability Outcome

Variable
LnY 33.12683 0.000000 Cointegration
Note: The asymptotic critical value bounds are min F=2.86 & Max F=4.01 at 5% (Table C1
iii. unrestricted intercept and no trend, Pesaran et. al. (2001).

Cointegration, however, cannot tell us the direction of Granger Causality

between the variables as to which is leading and which variable is lagging (i.e. which
variable is exogenous and which variable is endogenous). For discerning the
endogeneity and exogeneity of the variables of the variables, we can apply Vector
Error Correction Model (VECM). However, in absence of conintegrating relationship
among variables, we estimate Vector Autoregressive Model (VAR) by the exclusion
of the error-correction term for Granger Causality with short-term interactive
relationship (Granger 1988).

Third, the estimates of equation (5), the VECM are reported as follows
(Table-5).
 25

TABLE 5. ESTIMATING EQUATION (5) FOR VECTOR ERROR

CORRECTION MODEL (2, 1)

Variable Coefficient Std. Error t-Statistic Prob.

Constant 0.097691 0.018878 5.174922 0.0000
ECM t-1 -0.243625 0.070237 -3.468613 0.0017
∆LnY(-1) 0.157078 0.106239 1.478538 0.1504
∆LnY(-2) -0.234394 0.064965 -3.608019 0.0012
∆ LnBPSC 0.171165 0.069272 2.470907 0.0198
∆ LnBPSC(-1) 0.008942 0.070196 0.127388 0.8995
R-squared 0.688446
Prob (F-statistic) 0.000002 Durbin-Watson stat 1.647135

As revealed in Table-5, the coefficient of the error-correction term (ECMt–1)

has the expected negative sign with statistical significance. It implies that deviation
of the variables (represented by the error correction term) has a significant feedback
effect on economic growth that bears the burden of short-run adjustment to bring
about the long-term equilibrium. Additionally, the error correction model also helps
us distinguish between the short-term and long term Granger causality. The error
correction terms stands for the long-term relations among the variables. The
coefficient of ECMt–1 is equal to -0.243625 with statistical significance implying that
profound long-term relations exist between LnBPSC and LnY, and deviation to bring
about the long-term equilibrium is corrected by 24.36% over each year. It supports
that bank’s credit to the private sector is positively and robustly contributing to the
economic development of the country. In addition, contemporaneous change in
LnBPSC has profound positive short-run feedback effects to the economic growth of
the country as reflected in its coefficient and associated t-values. On the other hand,
lagged change of LnBPSC has also short-term positive but subdued effect to the
economic growth as it is evident from its coefficients and insignificant associated t-
value. The results of the ‘F’ tests i.e. 12.37442 also support the joint significance of
the lags of each of the differenced variables. The DW-value at 1.647135 indicates a
mild positive autocorrelation which is likely to be inherent in time series data. The
adjusted-R2 shows that about 69 per cent of Bangladesh’s economic growth is
explained by bank financing to the private sector. The optimum lag-lengths are
determined by the AIC (1969).

TABLE 6. ESTIMATING EQUATION (7) FOR VECTOR AUTOREGRESSIVE

MODEL (2, 4)
Variable Coefficient Std. Error t-Statistic Prob.
C 0.002200 0.001230 1.788644 0.0777
∆ LnY (-1) 0.949650 0.114562 8.289369 0.0000
∆ LnY (-2) -0.041741 0.115147 -0.362501 0.7180
∆ (LnCAPI) -0.002212 0.003000 -0.737486 0.4631
∆ LnCAPI(-1) 0.003312 0.003911 0.846692 0.3998
∆ LnCAPI (-2) -0.000146 0.003911 -0.037214 0.9704
∆ LnCAPI -3) 3.64E-15 0.003891 9.35E-13 1.0000
∆ LnCAPI (-4) 0.001577 0.003312 0.476137 0.6353

Adjusted R-squared 0.812567

Prob (F-statistic) 0.000000 Durbin-Watson stat 1.998560
 26

As long-run relationship does not exist between LnY and LnCAPI according to
prior estimation of cointegrating relationship between these two variables, VAR model is
estimated. Table-6 reveals that subdued net positive short term effect is noticeable from
LnCAPI to LnY as the respective sum of the lagged coefficients of variables is positive.
This is supported by F-statistics although associated t-value of coefficients of each
variable is statistically insignificant. The DW-value at 1.998560 indicates near absence of
autocorrelation. Finally, as stated above, in case of ARDL regression, long run
association has been shown in Table-7 and ECM version of this model shown in table-8.

TABLE 7. ARDL LONG RUN ESTIMATION OF LnY

Variables Co-efficient Std. Error t-statistics Prob.
c 21.42657 0.092986 230.4269 0.00000
LnBPSC 0.657227 0.010445 62.92389 0.00000
LNCAPI -0.006423 0.008517 -0.754067 0.4529

Table-7 confirms that the long run positive association between banks’ private
sector credit and economic growth of the country. But similar inference cannot be made
for relationship between equity market and economic development, as co-efficient shows
negative sign with insignificant t-value, which is absurd.

TABLE 8. ESTIMATING EQUATION 6 FOR ARDL (2, 1, 4) VECTOR ERROR

CORRECTION ESTIMATES
Variable Coefficient Std. Error t-Statistic Prob.
Constant 0.002019 0.001258 1.605233 0.1129
ECM t-1 -0.015350 0.009616 -1.596355 0.1149
∆ LnY (-1) 0.923078 0.119601 7.717992 0.0000
∆ LnY (-2) -0.007024 0.120674 -0.058206 0.9538
∆ LnBPSC 0.091808 0.028993 3.166547 0.0023
∆ LnBPSC) (-1) -0.077232 0.037809 -2.042672 0.0449
∆ LnCAPI -0.001513 0.003090 -0.489770 0.6258
∆ LnCAPI (-1) 0.001710 0.003841 0.445141 0.6576
∆ LnCAPI) (-2) -9.20E-05 0.004172 -0.022050 0.9825
∆ LnCAPI (-3) -0.000145 0.004372 -0.033122 0.9737
∆ LnCAPI (-4) -0.000833 0.003728 -0.223348 0.8239

Adjusted R-squared 0.823671

Prob (F-statistic) 0.000000 Durbin-Watson stat 1.912661

Table-8 reports that the estimated coefficients of error correction term are
negative, but statistically insignificant. It means that long run causal flows from both
the banks’ private sector credit and stock market development to Bangladesh
economic growth have been evident. However, this flow of relationship is weak as t
value is less than 2 meaning that Bangladeshi financial system comprising of both
banking sector and stock market jointly is not still a strong promoter of its economic
growth, although banking development is alone robustly associated with economic
development as we have seen earlier. But in case of short term interactive
relationship, it is evident that bank finance to private sector in contemporaneous level
shows a robust short-term positive causal effect to economic growth. The adjusted-R2
discloses a significant explanatory power of the model. The F-statistic is also quite
significant. The DW value shows near no-autocorrelation.
 27

CONCLUSIONS AND REMARKS

The findings from the empirical analysis indicate that bank’s credit to the private
sector is both positively and robustly contributing to the economic development of
the country. Besides long-term relationship, contemporaneous change in bank’s
credit to the private sector has profound positive short-run feedback effects to the
economic growth. It confirms that banks play a critical role in economic development
of the country. In the 1990s, banks have undergone massive reforms in line with
market based liberalization and continued with new and innovative reforms till to
date. These made this unit of the financial sector indispensable for the economic
growth of the country. This is substantiated by the overwhelming bank finance to
GDP ratio (68.75 per cent) compared to equity finance to GDP ratio (0.17 per cent)
prevailing in Bangladesh. However, a lot has still to be done to increase the
percentage of bank finance to GDP and most importantly to channelize more bank
finance for long term investment. On the other hand, in assessing impact of stock
market development on economic growth, no significant long-run relationship is
found between stock market development and economic growth in Bangladesh.
However, net positive short-term effect of stock market development on economic
growth is found as sum of lagged coefficients of the independent variable is positive
with significant F test. However, individual variables are not significant as evident
from associated t statistics. This is hardly surprising as the share market of the
country is now traversing an experimental period and this market is not generally
used by the business houses to generate their capital. In the final step, when both
bank and equity market jointly enter the model to find out the relationship of both
variables with economic growth, a long-run relationship has been evident without
statistical significance. But weak causal flows from both the banks’ private sector
credit and stock market development to Bangladesh economic growth have been
evident. This evidence could be interpreted to mean that Bangladeshi financial
system comprising of both banking sector and stock market jointly is not still a strong
promoter of its economic growth, although banking development alone is robustly
associated with economic development. It is mentionable that stock market could
hinder economic growth as in Taiwan. This happens when the stock market becomes
mostly a speculation instrument resulting in booms and busts that destabilize the
economy.
Bangladesh therefore needs to enhance the efficiency of banks for
increasing credit to the private sector as well as private sector investment as
proportions to GDP. An immediate initiative is also required to make both equity and
debt market as regular sources of finance for the real economy. In this perspective,
ensuring smooth operation of primary and secondary market, increasing financial
literacy among investors, minimizing volatility of the market, expanding issuer base,
creating both individual and institutional investors, enhancing efficiency of the
brokerage house, adding innovative financial services, initiating knowledge based
trading, rationalizing cost of generating funds and costs of funds, lessening
formalities involved and required documents, introducing shelf registration system,
creating more professional trustee and ensuring authenticate credit rating from the
rating agencies are required to be ensured.

ENDNOTES
1
For subsequent use in the vector error-correction version of ARDL model, the error-
correction term (𝐸𝐶𝑀𝑡−1 ) is obtained from the following equation:
 28

𝑝1 𝑞1
𝐸𝐶𝑀𝑡−1 = 𝑙𝑛𝑌𝑡 − (𝑎̂0 + ∑𝑖=1 𝑎̂1 𝑙𝑛𝑌𝑡−𝑖 + ∑𝑖=0 𝑎̂2 𝑙𝑛𝐵𝑃𝑆𝐶𝑡−𝑖 +
∑𝑞2 𝑎
̂
𝑖=0 3 𝑙𝑛𝐶𝐴𝑃𝐼𝑡−𝑖 )
2
Results are not reported in the paper but could be supplied on request.

REFERENCES

Akaike, H 1969, ‘Fitting Autoregression for Prediction’, Annals of the

Institute of Mathematical Statistics, vol. 21, pp. 243-247.
Antonios, A 2010, ‘Stock Market and Economic Growth: An Empirical
Analysis for Germany’, Business and Economics Journal, vol. 2010, BEJ-1.
Arestis, P, Demetriades, P and Luintel, KB 2001, ‘Financial Development
and Economic Growth: the Role of Stock Markets’, Journal of Money Credit and
Banking, vol. 33, pp. 16–41.
Asante, S, Agyapong, D and Adam, AM 2011, ‘Bank Competition, Stock
Market and Economic Growth in Ghana’, vol. 2, pp. 33-41.
Atje, R and Jovanovic, B 1993, ‘Stock Markets Development’, European
Economic Review, vol. 37, pp. 632-640.
Beck, T and Levine, R 2004, ‘Stock Markets, Banks and Growth: Panel
Evidence’, Journal of Banking and Finance, vol. 28, pp. 423-442.
Bencivenga, V and Smith, B 1993, ‘Some Consequences of Credit
Rationing in An Endogenous Growth Model’, Journal of Economic Dynamics and
Control, vol. 17, pp. 97–122.
Blackburn, K, Bose, N and Capasso, S 2005, ‘Financial Development,
Financing Choice and Economic Growth’, Review of Development Economics, vol.
9(2), pp. 135-149.
Boyd, J and Smith, B 1998, ‘The Co-evolution of the Real and Financial
Sectors in the Growth Process’, The World Bank Economic Review, vol. 10, pp. 371-
396.
Dickey, DA and Fuller, WA 1981, ‘Likelihood Ratio Statistics for
Autoregressive Time Series With A Unit Root’, Econometrica, vol. 49, pp. 1057-
1072.
Engle, R and Granger, CWJ 1987, ‘Co-integration and Error-Correction:
Representation, Estimation, and Testing’, Econometrica, vol. 35, pp. 315-329.
European Banking Federation (2012), European Banking Sector Facts and
Figures October 2012, Brussels: European Banking Federation.
Fuller, WA 1996, Introduction to Statistical Time Series, New York: John
Wiley and Sons.
Granger, CWJ 1988, ‘Some Recent Developments in a Concept of
Causality’, Journal of Econometrics, vol. 39, pp. 199-211.
Greenspan, A 2000, ‘Global Challenges’, Remarks at The Financial Crisis
Conference, Council on Foreign Relations, New York, 12 July.
Greenwood, J and Bruce S 1997, ‘Financial Markets in Development and
the Development of Financial Markets’, Journal of Economic Dynamics and Control,
vol. 21, pp. 145-182.
Herring, R and Chatusripitak N 2001, ‘The Case of Missing Market: The
Bond Market and Why It Matters for Financial Development’, Wharton Financial
Institutions Centre working paper, University of Pennsylvania, pp. 1-61.
Hondroyiannis, G, Lolos, S and Papapetrou E 2005, ‘Financial Markets and
Economic Growth in Greece’, Journal of International Financial Markets, Institutions
and Money, vol. 15, pp. 173-188.
 29

Islam, MA 2012, ‘Financial Sector Reforms’, Bank Prikrama, vol. XXXVII,

Special Issue (Nos. 3 & 4), September & December 2012, pp. 33-48.
Johansen, S and Juselius K 1990, ‘Maximum Likelihood Estimation and
Inference on Cointegration with Applications to The Demand for Money’, Oxford
Bulletin of Statistics, pp. 169-210.
Johansen, S 1991, ‘Estimation and Hypothesis Testing of Cointegration
Vectors in Gaussian Autoregressive Models’, Econometrica, vol. 59, pp. 169-210.
Johansen, S and Juselius K 1992, ‘Testing Structural Hypothesis in
Multivariate Cointegration Analysis of The PPP and VIP for U.K.’, Journal of
Econometrics, pp. 211-244.
Kwiatkowski, D, Phillips, PCB, Schmidt, P and Shin, Y 1992, ‘Testing the
Null Hypothesis of Stationarity Against the Alternative of a Unit Root’, Journal of
Econometrics, Vol. 54, pp. 159-178.
Levine, R 2002, ‘Bank-based or Market-Based Financial Systems: Which is
Better?’ Journal of Finance Intermed, vol. 11, pp. 398–428.
Levine, R 1997, ‘Financial Development and Economic Growth: Views and
Agenda’, Journal of Economic Literature, vol. 35, pp. 688–726.
Levine, R and Zervos, S 1996, ‘Stock Market Development and Long-Run
Growth’, World Bank Economic Review, vol. 10 (2), pp. 323-339.
Levine, R and Zervos, S 1998, ‘Stock Markets, Banks, and Economic
Growth’, American Economic Review, vol. 88, pp. 537-558.
Levine, R and Zervos, S 1993, ‘What We Have Learned about Policy and
Growth from Cross-Country Regressions?’ American Economic Review, vol. 83(2),
pp. 426-430.
Mackinnon, JG 1996, ‘Numerical Distribution Functions for Unit Root and
Conintegration Tests’, Journal of Applied Econometrics, vol. 11, pp. 601-618.
Pagano, M 1993, ‘Financial Markets and Growth: An Overview’, European
Economic Review, vol. 37, pp. 613-622.
Park, JH and Gyutaeg, Oh 2006, Developing Asian Bond Markets Using
Securitization and Credit Guarantees, in Asia’s Debt Capital Markets, eds by Arner,
D, Park, Jh, Lejot, P and Liu, Q, Milken Institute, USA.
Pesaran, MH, Shin, Y and Smith RJ 2001, ‘Bounds testing approaches to the
analysis of level relationships’, Journal of Applied Econometrics, vol.16 (3), pp. 289-
326.
Pesaran, H and Shin, Y 1999, ‘An Autoregressive Distributed Lag Modeling
Approach to Cointegration Analysis’, In S. Strom (eds.) Econometrics and Economic
Theory in the 20th Century: The Ragnar Frisch Centennial Symposium Cambridge
University Press.
Rajan, RG and Zingles, L 1998, ‘Financial Dependence and Growth’, The
American Economic Review, vol. 88 (3), pp. 559-586.
Rousseau, PL and Wachtel, P 2000, ‘Equity Markets and Growth: Cross
Country Evidence on Timing and Outcomes’, Journal of Banking and Finance, vol.
24, pp. 1933–1957.
Schumpeter, JA 1912, ‘Theorie Der Wirtschaftlichen Entwicklung’,
Leipzig, Germany: Dunker & Humblot.
Tang, D 2006, ‘The Effect of Financial Development on Economic Growth:
Evidence from the APEC Countries, 1981-2000’, Applied Economics, vol. 38, pp.
1889-1904.