FACTORS INFLUENCING PROFITABILITY IN THE
BANGLADESHI COMMERCIAL BANKING SECTOR
Farhana Yasmin Liza
*
Department of Business Administration
Shanto-Marium University of Creative Technology, Dhaka, Bangladesh
Abstract
This study explores the determinants of Bangladesh-based commercial banks’
protability of commercial banks in an emerging economy, Bangladesh. For this study,
commercial banks’ annual report data [over the period 2013-2024] was used. Probable
determinants were shortlisted using empirical literature. The study uses return on asset
(ROA) and return on equity (ROE) as the dependent variable. Panel regressions were run
using xed effects model and random effects model. The ndings show a signicant positive
relationship between asset quality and bank performance, and a signicant negative
relationship between capital adequacy and performance across all baseline models.
Keywords: ROA, ROE, Asset Quality, Bank Performance, Capital Adequacy, Fixed Effect
Model, Random Effect Model.
Introduction
Banks are depository institutions that connect savers and users of funds. These
intermediaries bridge the gap between nal borrowers and lenders, enabling well-organized
allocation of funds in the economy. Entities with excess funds can provide them to
entrepreneurs and other economic units who need funds to take advantage of economically
and nancially feasible investment opportunities, earning a rational return. The presence
of nancial markets and nancial institutions make such transfer of nancial resources
possible. As a result, both borrowers and lenders are better off compared to scenarios
without nancial organizations and intermediaries.
Financial institutions are argued to play an important role in funding and investment
through a multidimensional process that involves numerous interconnected and
interdependent factors of different natures. Assessing the contribution of each factor
independently is difcult. The key purpose of nancial institutions, along with other non-
depository institutions, is to help distribute the country’s scarce capital among various
alternative investment areas. Thus, nancial markets serve a dual role: they provide
numerous types of investment funds and impose discipline on businesses that are inefcient
and fail to pursue protable income objectives. As commercial banks expand globally, it
* Corresponsing author: Farhana Yasmin Liza, Email: iza82bd@gmail.com
Independent Business Review, Vol 15, No. 1 | June 2026 2
has become essential for nancial regulators to safeguard the stability and sustainability of
the banking system. It is observed that nancial institutions, especially commercial banks,
can support our economy’s expansion if they are properly organized and directed.
In Bangladesh’s context, efforts need to be designed to nurture banking activities and
accelerate the economic development of the country. Despite its signicant merits, it is
also important not to overlook the problems within the nation’s crisis-oriented banking
structure, which requires appropriate guidelines for banking procedures to safeguard
effective use and monitoring of business funds.
Literature Review
This chapter attempts to provide insight into various literature related to banks’
asset quality, capital adequacy, and bank performance in both international and domestic
contexts. It has been observed that numerous studies have been conducted using different
methodologies across various countries and nancial environments. The ndings of these
studies vary signicantly, showing differences across countries regarding protability, asset
quality, and bank performance. Here the researcher has also identied several potential
factors that may inuence commercial banks’ protability, asset quality, capital adequacy,
and other related matters.
This chapter offers a thorough review of literature on bank asset quality, capital
adequacy, and overall performance in both international and domestic settings. Various
studies have been carried out using different methodologies, reecting the diversity of
nancial environments and regulatory frameworks across countries. Some researchers
have used primary data sources, while others have relied on secondary data, leading to
diverse ndings.
Mendes and Abreu(2002) proposed that asset quality of bank could act as a potential
rescue mechanism for banks in the future. In their examination of Nigerian banks, Al-
Tamimi and A-Mazrooei(2013) stressed the importance of continuously analyzing and
investigating factors that inuence bank performance. Their results indicated ongoing
underperformance in the banking sector, marked by high credit risk and an increasing
number of bad loans.
Different studies indirectly link capital adequacy and Nonperforming loans to nancial
stability. The study of Ombaba(2013) on kenya’s high NPL rates for nine years ,summary
that for growth of economy and nancial crises, stability of banking sector is a condition.
That suggests that it is the bank level concern to manage NPL at microeconomic level.
Banks of Nigerian have highlighted the essential need for continuous analysis and
investigation of factors affecting bank performance, especially given the persistent
underperformance characterized by high credit risk and growing volumes of bad loans.
Muhmad and Hashim (2015) recognized asset management as a major management
concern for banks, observing that global surveys regularly highlight asset management
quality as a common priority among bankers worldwide. Vighneswara(20150 found that
Factors Inuencing Protability in the Bangladeshi Commercial Banking Sector 3
asset quality substantially inuences a bank’s ranking, with ratings and assessments of
management closely linked to asset quality. Marshall (1998) reinforced this perspective,
discovering that high quality assets are a characteristic feature of top performing banks,
while poor asset quality frequently corresponds with lower bank ratings. Yixin(2005)
noted an inverse relationship between non performing assets (NPAs) and bank income,
highlighting the negative effect on overall bank performance. To protect investor interests,
they recommended prudent credit risk management and strong asset protection strategies.
According to Poudel(2013) bank capital requirements are established to ensure
institutions maintain sufcient capital levels to reduce risks related to asset growth,
especially when assets are both protable and inherently risk .This regulatory framework
seeks to protect nancial stability by balancing growth with careful risk management.
Ariyanti.,D.(2010),observed that quality of asset is measured by comparing the
classication of credit, including doubtful, substandard and bad loans, with the total credit
provided by the bank. This assessment offers understanding into the bank’s capacity to
manage its income generating assets. Ariyanti (2010) further noted that the main objective
of allocating bank funds to productive assets is to generate expected revenue, emphasizing
the signicance of effective asset management in achieving nancial performance.
Hashim and Muhmad (2015) performed an analysis of the CAMEL framework,
consisting of capital adequacy, managerial competence, earnings quality, and liquidity, for
both domestic and international banks operating in Malaysia from 2008 to 2012.
Researchers worldwide have used ratios of protability to risk and solvency to liquidity
to examine the success or failure of different nancial organizations. These scholars
included Abata, M. A. (2014) applied ten nancial metrics to evaluate the effectiveness
and efciency of banking institutions. These parameters included credit risk ,protability,
and liquidity risk Marshall (1999), highlighted the importance of considering many factors
when analysing a bank’s level of performance.
Eyup et al (2018) investigated whether bad debts affect Turkish banks’ protability.
They used regression analysis to examine a time series data set containing 180 observations.
The relationship between bad loans or nonperforming and protability, as measured by
ROA and ROE, was found to be negative. Return on asset and return on equity were both
low when asset quality was poor and bad loan or nonperforming loans were high. Eyup
et al (2018) analyzed IFRS compliant quarterly nancial reports from 55 Turkish banking
sector banks. This study period covered the rst quarter of 2006 to the third quarter of
2018. Asset quality was measured by the ratio of bad or non performing (follow up) loans
to total assets. Based on the panel regression study ndings, there was a signicant inverse
relationship between bad loan or nonperforming loans and the ROA/ROE ratio variable,
a measure of bank protability. A decrease in asset quality or a high number of bad loans
could lead to bank bankruptcy during an economic crisis.
Ogbulu & Eze(2017) discovered that over the past 22 years, both domestic and
international agencies had implemented regulations to assess asset quality in relation to its
Independent Business Review, Vol 15, No. 1 | June 2026 4
importance. Revan (2000) conducted research on bank leverage (debt) and its drivers in
the United Kingdom, where leverage was the dependent variable, while long term loans,
short term debts, and retained earnings were the independent factors. Revan (2000) showed
that debt (also called leverage) was negatively correlated with risky assets, long term debt,
and short term debt. The Al Mekla Study (2003) assessed indicators of capital adequacy,
along with their effects on nancial risk indicators, bank revenues, and bank valuations. The
banking industry is vulnerable to interest rate risk, liquidity risk, and credit risk. Al Maleeji
(2001) conducted research to create a model for evaluating Egyptian banks and to establish a
standard that included various elements needed to assess capital adequacy. The model of Al
Maleeji (2001) reected most of the risks that commercial banks faced generally and credit,
ination, liquidity, and market risks specically. Khreaiwsh et al. (2004) examined the
different factors that determined the level of security maintained by Jordanian commercial
banks over a certain time period (1994 – 2004). According to the research results, there was
a strong positive relationship between banking security and ROA and ROE.
Banks were crucial to the expansion and growth of a nation’s economy. The reason was
that there was a positive relationship between the condition of the banking industry and
economic growth (Levine and Beck 2005).Besides the Central Bank, governments, banking
organizations, and other nancial institutions, Management of bank was signicant in
understanding the correlation between capital adequate and nancial sector performance.
Without enough capital, nancial industry could not operate efciently.
There were various factors that contributed to having sufcient capital. Bevan, A. A.,
and Danbolt, J. (2000) and J. M. and Van Horne (2005) suggested that regulators could
address moral hazard by focusing on capital adequacy regulation. The regulation of capital
adequacy protected small depositors. Small depositors make up the majority of bank’s
funds, so this protection was very important.
Financial Sectors of Bangladesh
Bangladesh’s banking sector demonstrates the nation’s continuous commitment
to achieving nancial excellence. This overview highlights the compelling narrative of
banking performance in this dynamic country.The Bank of Bangladesh (BB) predicted
that the economy of Bangladesh would remain prosperous and stable in scal year 2024
provided these policies were upheld.It is noted that across all nancial sectors, the banking
sector’s contribution exceeds the combined contributions of other sectors when considering
total assets held relative to Gross Domestic Product(GDP).Commercial banks have had a
substantial inuence on the economy of Bangladesh throughout the years. They supply
both the public sector and private industry with investable capital to speed up the nation’s
economic development.
Indicators of major performance of Banks in Bangladesh
Performance of bank can be assessed through various theoretical and analytical
parameters. In this section, efforts have been made to highlight the key performance
indicators of banks functioning in Bangladesh.
40
30
20
10
0
-10
-20
-30
-40
2014
2015
2016
2017
2018
2019
Total
SCBs
SBs
PCBs
FCBs
2020
2021
2022(June)
1.2
1
0.8-
0.6-
0.4-
0.2-
0
12
3
4
5
67
8
9
10
11
12
Aggregate
Capital
Adequacy
Ratio
0.9
0.8
0.7
0.6
0.5-
0.4-
0.3
0.2
0.1-
1
2
3
4
5
6
7
8
9
10
11
12
ROA
12
ROE
10
2
0
1
2
3
4
5
6
7
8
6
10
11
12
ROE%
Independent Business Review, Vol 15, No. 1 | June 2026 8
amount of growth in 2022 as well, representing 9.37% of ROE. Except for SBs, all other
banks experienced positive protability growth.
Among the four categories of banks, specialized private banks (SPBs) recorded the
lowest ratio of nonperforming assets, reecting stronger risk management practices. In
contrast, stateowned commercial banks (SCBs) have long faced challenges in maintaining
sound loan portfolios due to issues such as weak oversight, political inuence, and limited
accountability. To address the high levels of nonperforming loans in SCBs, Bangladesh
Bank has introduced measures aimed at improving governance and operational efciency
Methodology
The nancial information used in this research consists of banking companies categorized
as State-Owned Commercial Banks and National Private Commercial Banks in Bangladesh
from 2012 to 2024, with both types having the following characteristics: they are not Islamic
banks and the researcher has complete annual reports from 2012 to 2024. Based on the
characteristics listed above, the number of banks that could fulll the criteria were 84 banks,
comprising 28 state-owned commercial banks and 64 national private banks. This research
concentrated only on Bangladesh-based commercial banks; therefore quite naturally, data
collection, interpretation of results. Research philosophy can be broadly divided into two
streams – positivism and phenomenology. There exist slight differences between these two
research philosophies according to three dimensions: axiology, ontology and epistemology.
Philosophy of research actually guides the researcher through multiple stages – null
hypothesis formation, data collection, interpretation of results. Research philosophy can be
broadly divided into two streams – positivism and phenomenology.
The researcher attempted to focus on the protability of commercial banks and that is
why researcher selected only non-Islamic commercial banks. The researcher has utilized
panel database. The research time period extended from 2012 to 2024.This panel database
had both cross-sectional and time-series variations. Here researcher has conducted a series
of regressions to determine the determinant factors of the protability of commercial banks.
The panel data-based regression in order to capture the cross-sectional variance and
time variance among the variables. Though it will be a short panel, it will be a balanced
panel as well. It is expected that the collection of all the necessary data has been collected.
Hausman test will be conducted to ascertain whether the researcher will go for a random
effects model or xed effects model. The panel regression will look like the following
equation.
ROA
it
= +
j
X
j
i,t
+
m
X
m
t
+
n
X
n
t
+ v
i,t,
……………………(1)
So, it is established through the panel regression that accounting protability of
commercial banks is dependent on three factors – bank specic factors, macro-economic
factors and industry structure. Bank specic factors are as follows: business size, capital
adequacy, asset quality, cost management and activity mix. In the above mentioned
regression, notation ROA
it
refers to the return on asset of a particular commercial bank
9
relevant for a particular year. X
j
ict ,
X
m
,t
, X
n
t
refers to the vectors of commercial bank
specic, macro-economy specic and industry specic determinants. V
i,t
refers to the
error part of the regression that covers for the effect due to the unobserved factors and the
idiosyncratic errors as well (Bryman and Bell, 2011). The dynamic version of the model
is as follows:
ROA
it
= +
j
X
j
i,t
+
m
X
m
t
+
n
X
n
t
+ v
i,t +
ROA
i,t-1
In the above mentioned model ROA
i,t-1
refers to the one-period lagged dependent
variable and refers to the speed of mean reversion whereby the adjustment co-efcient
can move in between 0 to 1. Since, accounting prot often does not t well with economic
prot angle, the researcher will also explore the economic prot dynamics of commercial
banking in Bangladesh. The generalized panel regression for understanding economic
prot dynamics will look like the following equation.
ROE
it
= +
j
X
j
it
+
m
X
m
,t
+
n
X
n
t
+ v
i,t …………………….(.2)
So, it is established through the panel regression that economic protability of a
commercial bank is dependent on three factors – bank specic factors, macro-economic
factors and industry factors. In the above-mentioned regression, notation RE
it
refers to the
residual income of a particular commercial bank and relevant for a particular year. X
j
ic,t
,
X
m
,t
, X
n
t
refers to the vectors of commercial bank specic, macro-economy specic and
industry specic determinants. v
i,t
refers to the error part of the regression that covers for
the effect due to the unobserved factors and the idiosyncratic errors as well. The dynamic
version of the model is as follows:
ROE
it
= +
j
X
j
it
+
m
X
m
t
+
n
X
n
t
+ v
i,t +
RE
i,t-1
In the above mentioned model RE
i,t-1
refers to the one-period lagged dependent variable
and refers to the speed of mean reversion whereby the adjustment co-efcient can move
in between 0 to 1. The researcher will also try to understand the market perception of
commercial bank’s protability through CAPE (cyclicality adjusted price-earning ratio).
The generalized panel regression for understanding CAPE dynamics will look like the
following equation.
CAPE
it
= +
j
X
j
it
+
m
X
m
,t
+
n
X
n
t
+ v
i,t ………………………..(3)
So, it is established through the panel regression that market-determined price-earning
multiple of a commercial bank is dependent on three factors – bank specic factors, macro-
economic factors and industry factors. In the above-mentioned regression, notation RE
it
refers to the residual income of a particular commercial bank and relevant for a particular
year. X
j
ic,t ,
X
m
,t
, X
n
t
refers to the vectors of commercial bank specic, macro-economy
specic and industry specic determinants. v
i,t
refers to the error part of the regression that
covers for the effect due to the unobserved factors and the idiosyncratic errors as well. The
dynamic version of the model is as follows:
CAPE
it
= +
j
X
j
it
+
m
X
m
t
+
n
X
n
t
+ v
i,t +
CAPE
i,t-1
Factors Inuencing Protability in the Bangladeshi Commercial Banking Sector
Independent Business Review, Vol 15, No. 1 | June 2026 10
In the above mentioned model CAPE
i,t-1
refers to the one-period lagged dependent
variable and refers to the speed of mean reversion whereby the adjustment co-efcient can
move in between 0 to 1.
Now the researcher will delineate the operational denition of the independent and
dependent variable.
Table 3. Operational denition independent variables
Independent Variable Operational denition Expected Sign
Business Size Natural logarithm of commercial bank’s
total asset
Positive/negative
Capital Adequacy Tier 1 and tier 2 capital/risk weighted asset Positive/negative
Activity Mix Total non-operating income /total income Positive
Asset Quality Substandard, doubtful and bad and loss loan/
total loan
Positive
Net Interest Margin (Interest income/ Interest expense) average
earning assets
Positive
Market Structure HH (Herndahl- Hirschman) Index Positive/negative
Cost Management Natural logarithm of commercial bank’s
overhead
Positive
Term Structure of Interest
Rate
Difference of yield spread of 10years and 5
years treasury bonds
Negative
Ination Rate % change in year-to-year CPI level Positive/negative
Macro-Economic Growth
Rate
% Change in annual GDP of Bangladesh Negative
Source: Researcher
Table 4. Operational denition dependent variables
Dependent variable Operational denition
ROA Net income/total asset
ROE Net income – (equity*cost of equity)
CAPE Price/(10-year moving average of EPS)
Source: Researcher
Given, these pre-conditions in a single factor model, we can estimate beta estimators
using the following way.
11
There is no other way to simply the expression any more. In order to simply this
expression further, the researcher is going to use expectation operator. By taking expectation
operator in both the sides:
Since we are running a CLRM, the mean independence condition must have been
fullled. And Mean independence: E[ui│Xi]= 0; E(uixi) = 0
This single factor regression model-based proof of unbiasedness can be extended
in a multi-factor model and the research area of interest revolves around a multi-factor
model [where commercial banks’ protability is dependent on a number of factors]
(Woolridge,2015). So, the above-mentioned proof of unbiasedness is a legit framework in
a multi-factor model.
Given, these pre-conditions in a single factor model, we can estimate beta estimators
using the following way.
Factors Inuencing Protability in the Bangladeshi Commercial Banking Sector
Independent Business Review, Vol 15, No. 1 | June 2026 12
There is no other way to simply the expression any more. In order to simply this
expression further, the researcher is going to use plim operator. By taking plim operator in
both the sides:
The consistency proof in a single-factor regression model can be extended to a multi-
factor framework. Since commercial banks’ protability depends on several variables, this
approach offers a valid basis for analysis in multi-factor models (Woolridge, 2015).
Findings and Discussion
To estimate a determinant model for bank protability, the researcher has conducted
a number of panel regressions. Before conducting panel regressions, it was necessary to
determine whether the estimated effects would be a random-effects or a xed effects model.
A xed-effects model assumes that there exists cross-sectional heterogeneity in the dataset
and it can be handled within the given structure. Conversely, in case of random-effects
model, it is assumed that there exists no cross-sectional heterogeneity – a very restrictive
and impractical assumption. By conducting the Wu-Durbin-Hausman test, one can also
determine whether a xed-effects model or a random effects model better suits the purpose.
The null and alternative hypothesis related to the Wu-Durbin-Hausman test are as follows:
Null hypothesis: The given panel regression can be better explained by random effects
13
framework. Alternative hypothesis: The given panel regression can be better explained by
xed effects framework.
Based on a 95% condence level, these hypotheses were tested. For all the regression
models, the preferred model is xed effects model. Panel regressions are conducted
within the OLS framework and that is why panel regressions must meet certain OLS pre-
conditions.
The estimated relationship between the factor and independent variable is linear. The
data is collected through random sampling. The covariance (hence correlation) between
independent variables and error is zero.There exists no perfect multicollinearity among the
pairs of independent variables.
ROA
i,t
= α +
i
Firm-specic factors
i,t
+ β
7
Industry-specic factor +
j
Macro-economic
factors
t
+ β
11
Liquidity proxy
i,t
+ β
12
Managerial efciency proxy
i,t
+ β
13
Labor market
proxy
i,t
+a
i
+ ∆
t
+ ε
i,t
In the above-mentioned model, ROA is the dependent variable and the researcher has
tried to estimate a determinant model for ROA. ε
i,t
refers to the disturbance element and
it can be decomposed into three sub-components such as – cross-sectional, time invariant
heterogeneity (ⱷ
i
), time effect (δ
t
) and noise (γ
i,t
).
Table 5. Baseline Regression result
Variables ROA
Size 0.00703**
2.069
Capital Adequacy -0.0568***
(-7.454)
Activity Mix 0.0145**
(2.057)
Asset Quality 0.000185
(0.558)
NIM 0.611***
(19.87)
HHI 0.177
(0.397)
Cost Management 0.00537*
(1.985)
Interest Rate 0.186
(0.474)
Ination -0.00790
(-1.111)
Factors Inuencing Protability in the Bangladeshi Commercial Banking Sector
Independent Business Review, Vol 15, No. 1 | June 2026 14
GDP Growth Rate 0.0895
(1.398)
Constant 0.0375
(0.673)
Observations 245
R-squared 0.747
Number of Firm Code 27
Firm FE Yes
Firm Cluster No
seasonal adjustment No
R2 within 0.747
R2 overall 0.578
R2 between 0.0800
F-stat 67.97
Prob > F 0
t-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Complied by: The researcher
Interpretation of the second baseline regression:
a. Theoretical framework: The researcher has tried to understand the relationship
between economic prot and various rm-specic and macro-economic variables
through regression.
b. Regression magnitude and sign : The individual, factor-specic test of hypothesis
was conducted by t-test. The null hypothesis [Ho: β
1
=0] is rejected at 5% level of
signicance. It was found that activity mix, business size, cost management, interest
rate, GDP growth rate, asset quality, net interest margin positively inuenced the
economic prot of commercial banks. It was also evident that capital adequacy and
ination rate negatively inuenced the economic prot of commercial banks.
c. R-square: It represents to the explanatory power of the model. It is around 13%; so
13% of the changes in the dependent variable (economic prot) can be explained by
the combined changes in the independent variable.
d. Model signicance: It was hypothesized that the estimated model is not signicant;
so it was assumed that Ho: β
1
=β
2
= …= β
10
=0. This joint test of hypothesis was
conducted through F-test and at a 5% level of signicance this joint hypothesis was
rejected. So, at least one of the factors is different from zero and it is a signicant
model to predict economic prot.
15
The baseline regression model can be represented through the following equation.
ROE
i,t
= α +
i
Firm-specic factors
i,t
+ β
7
Industry-specic factor +
j
Macro-economic
factors
t
+ β
11
Liquidity proxy
i,t
+ β
12
Managerial efciency proxy
i,t
+ β
13
Labor market
proxy
i,t
+a
i
+ ∆
t
+ ε
i,t
CAPE
i,t
= α +
i
Firm-specic factors
i,t
+ β
7
Industry-specic factor +
j
Macro-economic
factors
t
+ β
11
Liquidity proxy
i,t
+ β
12
Managerial efciency proxy
i,t
+ β
13
Labor market
proxy
i,t
+a
i
+ ∆
t
+ ε
i,t
In the above-mentioned model, CAPE is the dependent variable and the researcher has
tried to estimate a determinant model for CAPE. ε
i,t
refers to the disturbance element and
it can be decomposed into three sub-components such as – cross-sectional, time invariant
heterogeneity (ⱷ
i
), time effect (δ
t
) and noise (γ
i,t
).
Table 6. Baseline Regression result
Variables CAPE
Size 2.240
0.626
Capital Adequacy -2.7821
(-0.8681)
Activity Mix 9.99*
2.0032
Asset Quality 0.102
(0.496)
NIM 8.0916
(0.301)
HHI 99.6
(0.276)
Cost Management 2.429
(1.319)
Interest Rate 214.6
(0.323)
Ination -4.852
(-0.672)
GDP Growth Rate 23.72
(0.382)
Constant 18.32
(0.381)
Observations 245
Factors Inuencing Protability in the Bangladeshi Commercial Banking Sector
Independent Business Review, Vol 15, No. 1 | June 2026 16
R-squared 0.027
Number of Firm Code 24
Firm Fixed effect No
Firm Cluster Yes
seasonal adjustment No
R2 within 0.03001
R2 overall 0.0358
R2 between 0.141
F-stat 0.822
Prob > F 0.004
t-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
*** p<0.01, ** p<0.05, * p<0.1
Complied by: The researcher
Interpretation of the baseline regression:
e. Structural Foundation: Through regression, ROE and various rm-specic and
macro-economic variables has relationship which was tried to understand
f. Nature and scale of estimated effects: t-test conducted the individual, factor-
specic test of hypothesis. The null hypothesis [Ho: β1 =0] is rejected at 5% level
of signicance. Activity mix, business size, net interest margin ,cost management,
interest rate, GDP growth rate, asset quality, positively inuenced the ROE of
commercial banks. It was also evident that capital adequacy and ination rate
negatively inuenced the ROE of commercial banks.
g. Regression model: It refers to the explanatory power of the model. It is around 3%;
so, 3% of the changes in the dependent variable (ROE) can be discussed by the
changes in the independent variable.
h. Signicance Model: It was hypothesized that the model is not signicant; so, it
was assumed that Ho: β1 =β2 = …= β10=0. This combined test of hypothesis
was established through F-test and at a 5% level of signicance this combined
hypothesis was rejected. So, the factor is different from zero and ROE was predicted
to be signicant in this model . After reviewing the literature, there is a signicant
number of bank-specic and macro-factors that have an impact on commercial
bank’s protability. The study present the baseline regression results where the
partial effect of various factor variables on banks’ protability was assessed along
with these two omitted variables.
ROA
i,t
= α +
i
Firm-specic factors
i,t
+ β
7
Industry-specic factor +
j
Macro-economic
17
factors
t
+ β
11
Liquidity proxy
i,t
+ β
12
Managerial efciency proxy
i,t
+ β
13
Labor market
proxy
i,t
+a
i
+ ∆
t
+ ε
i,t
ROE
i,t
= α +
i
Firm-specic factors
i,t
+ β
7
Industry-specic factor +
j
Macro- economic
factors
t
+ β
11
Liquidity proxy
i,t
+ β
12
Managerial efciency proxy
i,t
+ β
13
Labor market
proxy
i,t
+a
i
+ ∆
t
+ ε
i,t
CAPE
i,t
= α +
i
Firm-specic factors
i,t
+ β
7
Industry-specic factor +
j
Macro-economic
factors
t
+ β
11
Liquidity proxy
i,t
+ β
12
Managerial efciency proxy
i,t
+ β
13
Labor market
proxy
i,t
+a
i
+ ∆
t
+ ε
i,t
Table 7
VARIABLES Model 1 Model 2 Model 3
ROA ROE CAPE
Size 0.0069 0.0053 0.006
3.00867 0.6572 0.95106
Capital Adequacy -.990332* -6.18 -4.72651*
-2.0092 -3.714 -3.2811
Activity Mix 0.03135 1.34E+09 16.457
2.89861 2.01252 2.2692
Asset Quality 3.025** 3.989* 6.1362**
.79794 .25334 .51876
Net Interest Margin 0.87373 3.22E+09 1.36996
2.4131 1.62498 0.40431
HHI -0.25311 -8.43+10 -3.096
-0.56914 -3.05068 -0.36156
Cost Management 0.006249 2.302 3.31299
2.69555 0.16006 1.72789
Interest Rate -0.26026 -2.61010 -0.126
-0.60632 -0.97944 -0.42313
Ination -0.01101 -4.8908** -5.04612**
-1.58873 -1.13526 -0.88032**
GDP Growth Rate 0.127699* 4.6309** 2.4532*
1.85614 .08438 -.50042
Liquidity -.564764 -3.005 -2.06048
-1.31571 -2.1253** -0.203**
Managerial efciency 0.02389 0.987 2.32214
Factors Inuencing Protability in the Bangladeshi Commercial Banking Sector
Independent Business Review, Vol 15, No. 1 | June 2026 18
3.44754 2.46351 0.3255
Labor efciency 0.276107 0.008 3.45254
4.02782 2.3531 0.24302
Constant -2.0980 -3.9890 -2.520
Observations 245 245 245
R-squared 0.342 0.219 0.251
Number of Firm Code 29 29 29
Firm xed effects Yes Yes Yes
Firm Cluster No No No
seasonal adjustment Yes Yes Yes
R2 within 0.017 0.052 0.034
R2 overall 0.076 0.015 0.032
R2 between 0.231 0.256 0.312
F-stat 6.020 4.045 5.067
Prob > F 0.005 0.006 0.024
t-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Compiled by: The researchers
Correlation Matrix
Table:8 shows that return on asset(ROA) is negatively correlated with net interest
margin(NIM),return on equity (ROE),capital adequacy ,market structure and asset
quality(AQ),activity mix, nonperforming loan(NPL),ination and negatively correlated with
cost management, total asset, leverage ,loan loss provision, and GDP, Return on asset and return
on equity are positively correlated with NIM ,ROE, NLP and GDP. From table 8 it is strongly
represent the proxies of protability used in this study are negatively related to the proxy of cost
management. So, this can be mentioned from the relationship of the variables that a bank uses
NPL to pursue their interest directly or indirectly, can affect the nancial efciency negatively.
Table 8. Represents the correlation coefcient of the variables used in the study of
commercial banks from2022 to2024.
Variables 1 2 3 4 5 6 7 8 9 10
1. Business size 1
2. Capital adequacy 0.47 1
3. Activity mix 0.43 0.87 1
19
4. Asset quality -0.1 -0.2 -0.1 1
5. Net interest margin 0.4 0.88 0.78 0.88 1
6. Market structure 0.23 0.45 0.67 -0.6 0.9 1
7. Cost Management -0.1 -0.1 -0.3 -0.3 -0.4 0.24 1
8. Term structure of
Interest
-0.2 0.42 0.56 0.45 0.6 0.45 0.58 1
9. Ination -0.1 -0.2 -0.3 0.67 0.9 0.38 0.78 0.7 1
10. Macro-economic
growth rate
0.02 0.07 0.09 0.34 0.5 0.27 0.53 0.3 -1 1
Here the main objective of the analysis will be to understand how commercial
bank operating in Bangladesh makes prot and how to make that prot sustainable
in the longer run. For that reason, the researcher will conduct a positivism driven
research study; where secondary data will be used to understand the protability
dynamics. Systematic panel data regression will be the data analysis tool.
Table 9. Out of sample predictability
Machine learning tools Accuracy score Training-testing split Target variable
Regression Logistic 87% 75%-30% ROA[transformed]
Forrest Random 76% 75%-20% ROA[transformed]
Machine Support vector 65% 75%-20% ROA[transformed]
Decision-tree 73% 72%-25% ROA[transformed]
Compiled by: The Researchers
The baseline model has achieved memorable accurate scores in predicting the protable
direction in the out-of-the-sample contexts. Here the analysis of three xed-effects
regression models represent relationship between dependent variables and independents
variables, with different degrees of value and signicant model. This factor-specic test
of hypothesis was conducted by t-test. The null hypothesis [Ho: β
1
=0] is rejected at 5%
level of signicance. Bank performance and asset quality is related positively which was
evident and this regression model was signicant. On the other hand, banks’ performance
and capital adequacy performance is negatively related which was observed and in case
of all the baseline regression models, the regression coefcient is signicant, once the
omitted variables were introduced in the models. Overall baseline models are statistically
signicant which is mentioned in this study.
Conclusion
Commercial banks play a crucial role in fostering a nation’s economic growth by
channeling capital into productive investments. Actual rising competition, increasing risk
exposure, and rapid technological change, these institutions must strengthen their efciency
Factors Inuencing Protability in the Bangladeshi Commercial Banking Sector
Independent Business Review, Vol 15, No. 1 | June 2026 20
to remain resilient and effective. It is difcult and challenging task to manage a bank. The
challenge has become greater because of the recent structural changes in the environment,
economy, technology, competition, nancial liberalization and relaxation or inclusion of
new rules and regulation in which a bank operates. Hence, focus is given to the managerial
techniques that help optimize the trade-off between risk and return of a bank within the
constraint imposed by its environment in managing its funds and investments. Bangladeshi
bank is observed to be alarming as maximum bank do not have any investment at all due
to economic problems and political unrest they are unwilling to invest. This is a crucial
factor that directly affecting protability as well. The sequence of this analysis suggests
that Bangladesh cannot adopt bank based economic structure due to some differences
in corporate practices in the economy of country.For this purpose proper regulation of
the banking operations is required to mitigate the current crisis ridden banking system of
Bangladesh. Moreover internal efciency and proper and efcient use of funds should be
ensured.
The substantial features and contribution of this research are provided below:
Based on the PCA results, it was concluded that ‘dimension reduction’ would not be an
appropriate approach to understanding the protability determinants of Bangladesh-based
commercial banks since the rst two principal components can together explain only 46%
of the total variation.
This research designed three regression models for the protability of Bangladesh-
based commercial banks. Commercial bank’s protability was through three perspectives:
accounting prot (ROA),CAPE(residual income), and the market’s perception of prot
(ROE). It was clear that some bank-specic factors, macroeconomic, and commercial
industry-specic variables were involved with commercial banks’ protability.
The regression results showed that factors like activity mix, business size, GDP growth
rate, cost management, interest rate, net interest margin negatively impacted commercial
banks’ protability. On the other hand, capital adequacy, asset quality and ination rate
were found to positively affect accounting protability.
Research ndings indicate that variables such as organizational scale, diversication of
operations, expense control, lending rates, national economic expansion, asset quality, and
net interest yield contribute positively to the nancial performance of commercial banks.
Conversely, capital sufciency and rising price levels were shown to exert a detrimental
inuence on economic returns.
This research conrmed that the estimated relationships remain the same across
different model specications, as the movement of the coefcients and their statistical
signicance are not changed when using a random effects approach. Furthermore, the
ndings proved consistent in out-of-sample testing, using a 60%-80% training-to-testing
partition. Furthermore, results indicated that asset quality continued to show a negative
relationship with bank performance, while capital adequacy maintained a positive
relationship with both effects remaining statistically signicant across all baseline models
21
under alternative denitions. No evidence of cross-sectional heterogeneity was detected in
the estimated results.
Reference
Abata, M. A. (2014) “Asset quality and bank performance: A study of commercial banks in Nigeria”,
Research Journal of Finance and Accounting, 5(18), 39-44.
Abreu M. and Mendes V. (2002),”Commercial bank interest margins and protability:evidence
from E.U countries”,Porto Working paper series” Research Gate.
Achou,T.F. and Tenguh, N. C. (2008),”Bank performance and credit risk management”. Master
Degree Project, Finance University of Skodve School of Technology and Society.
Al Maleeji and Hisjham. (2002). Development of accounting measurement for Judging Capital
adequacy in commercial banks, An empirical study, commerce collage scientic’. Journal, Issue
no.26, pp329-412
Al- Mikhla, Abdel Aziz (2004).”Analysis of banking capital adequacy According to international
standards, A case study of Yemeni Commercial banks”, unpublished PHD, National information
center, Yemin WWW.Yemin- – nic.info.
Al-Tamimi and Al-Mazrooei(2013),Banks’ risk management a comparison study of UAE national
and foreign banks,’International Journal of Academic Research in Economics and Management
Sciences,”,July2013,Vol.2,No,4 ISSN;2222-699.
Ariyanti.,D.(2010), “Macro-nancial link and monetary policy management in Indonesia,” Macro-
Financial Link and Monetary Policy Management, 73-94.
Beck, T., and Levine, R. (2004), “Stock markets, banks, and growth: Panel evidence”, Journal of
Banking & Finance, 28(3), 423-442.
Bevan, A. A., and Danbolt, J. (2000), “Dynamics in the determinants of capital structure in the UK.”
Khrawish, Al-Zoubi and Al-Abadi (2004), “Factors Affecting the level of Banking Hedging: Applied
Application,” Economics and Administration Journal, Volume 18, Issue Number 2,1425 AH, 2004
AD.
Muhmud,S.N.,andHashim,H.A.92015),:Using the camel framework in assessing bank performance
in Malaysia”, International Journal of Economics, Management and Accounting,23(1)
Marshall (1999), “Bank Capital Standards for Market Risk: A Welfare Analysis, Review of Finance,”
Volume 2, Issue 2, 1999, Pages 125–157, https://doi.org/10.1023
Nicholas K. S., Tobias O. and Maurice S.(2019), “Asset Quality As A Determinant Of Commercial
Banks Financial Performance in Kenya,” International Journal of Economics, Commerce and
Management, Vol. VII, Issue 2, February 2019, ISSN 2348 0386.
Ombaba, k.B.M. (2013),’Assessing the factors Contributing to Non-Performance Loans in
Kenyan Banks”,EuroJournal of Business and Management,ISSN2222-1905(Paper)ISSN2222-
28399Online),Vol.5,No,32,2013.
Pamungkas N.P. and Sutisnawati Y. (2022), “State-Owned Bank Health Level using the CAMELS
Factors Inuencing Protability in the Bangladeshi Commercial Banking Sector
Independent Business Review, Vol 15, No. 1 | June 2026 22
Method for the Period of 2015-2020,” JIKA, Vol 12 No 1: Desember 2022, DOI: https://doi.
org/10.34010/jika.v12i1.8304
Poudel(2013),”Non-performing Assets and Bank Protability of Nepalese Commercial
Banks”,Nepalese Journal of Business, Volume 3,Number3,July 2016.
Siahaan and Asandimitra (2006), “The Inuence of Liquidity and Asset Quality on Protability in
National Commercial Banks” (Study on the Indonesian Stock Exchange for the 2010-2014 Period),
BISMA (Businessand Management), 9(1), 1-12. https://journal.unesa.ac.id/index.php/bisma/index.
Van Horne J. C. and Wachowicz Jr. J. M. (2005), “Financial statement analysis”, Fundamentals of
Financial Management, (11th edition). India, Pearson, pp. 125-168
Vighneswara, S.(2015),Modeling Bank Asset Quality and Protability: An Empirical Assessment”,
Economics Discussion Papers No2015-27,Kiel Institute for the World Economy.
Yixin,H.(2005)”The Non-performing loans:some bank evidence”, Department of Economics
University of Birmington,Edgbaston Birmingham.
