Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.1, 2017
47
Determinants of Bank Liquidity: Empirical Evidence from Listed
Commercial Banks with SBP
Farooq Ahmad
Lecturer, Karakorum International University, Gilgit-Baltistan (Pakistan)
Nasir Rasool
Assistant Professor, Capital University of Science and Technology, Islamabad (Pakistan)
Abstract
This study empirically investigates the determinants of commercial banks liquidity; we took a sample size of 31
listed commercial banks with state bank of Pakistan from a population of 37 commercial banks. A convenience
sampling method is used to collect data for the period of 10 years, starting from 2005 up to 2014.The stock
approach method was used to measure the bank liquidity. The results of balance fixed effect model showed that
the independent variables like CAP and GDP have positive and significant impact on bank liquidity while NPL
and BS have statistically significant and negative impact on bank liquidity. Subsequently we found that ROE and
INF have statistically insignificant but positive relationship with bank liquidity. Moreover, commercial banks in
Pakistan should not only be focused about bank specific variables, but they must consider both the internal and
external factors together in developing strategies to improve the liquidity position of the banks. The results of
this study are important for credit manger, regulators and academician, in the sense, that they can facilitate
commercial banks in efficient resource allocation.
Keywords: Bank liquidity, Liquidity risk, Financial Institutions, State Bank of Pakistan (SBP), Bank for
international settlement (BIS)
Introduction
Commercial Banks are major players in the financial universe; this fact is proved by the financial crises in 2007-
08. The commercial banks perform the key role as transfer surplus funds from developed sectors to needy sectors
and in this way banks create a balance between surplus economic units and deficit business units and strengthen
the overall economic condition of a specific country. BCS, (2000) explains that bank liquidity is necessary for
banks’ daily routine operations to pay the claim of their short term depositors as well as short term business
obligations. If banks could not satisfy the claim of depositors then this will create the banks to face liquidity
shocks and ultimately, banks are going to bankruptcy or liquidation. . BIS, (2008) said that bank liquidity means
to satisfy the claim of depositors as they come due, without further undesirable losses. Diamond and Dybvig,
(1983) concluded that the main reasons that the banks are delicate; their role in transforming maturity and
providing assurance in respect of short term depositor’s that whenever they need their deposits, the bank will
satisfy their claims. This argument of bank fragile is supported by most recent researchers (Rauch et al. 2009).
Hence; liquidity is the key element of banks to safeguard against bankruptcy.
Basel 111, (2010) was published for the purpose to overcome the shortfall of Basel 11 regarding bank
liquidity. Basel report, (2010) clearly states that bank should maintain the liquidity coverage ratio which reflects
that reasonable level of liquid assets and must be fulfilling the liquidity provisions for a one month time period
under a rigorous state of liquidity stress. Basel 111 has highlighted the importance of holding liquid assets. In the
event of recession in a country, banks with more liquid assets have better survival chances than those banks with
less liquid assets. This will encouraged the banks to hold more liquid assets to control the economic downturn.
During the subprime crisis, large banks failed due to lack of liquidity even if they received extensively liquidity
support. After this crisis, the regulators start to make proposals to implement liquidity ratios in addition to capital
standard. Moreover, Ionica Munteanu, (2012) concluded that the lack of bank liquidity are caused by global
crisis as well as all negative events. The lender of the last resorts support to commercial banks regarding bank
liquidity, even with such far-reaching support, many financial institutions were declared bankrupt even they
were profitable due to liquidity mismanagement as in the case of Lehman Brothers in 2008.
Theoretical Background
The notion of bank liquidity has received substantial attention from both researchers and popular academics.
Various studies have been carried out to investigate the bank liquidity and its determinants. Keynes, (1936) has
presented Liquidity Preference Theory and recognized that three reasons on why people demand and prefer
liquidity. The transaction motive of holding cash means daily transactions of the company to keep the business
wheel turning. The precautionary motive reflects that a company must also keep liquidity for meeting unforeseen
or unexpected cash out flows. Speculative reason refers to business units prefers liquidity to take advantage of
special investment opportunities which will result increase the profit of banks.
Wang, (2002) says that banks make sure the availability of liquidity in the economy by accepting
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.1, 2017
48
deposits, liquid liabilities and then advancing long term loans to economy against demand deposits and keep
illiquid assets. Banks face transformation risks and ultimately, bank runs on deposits occurred. Drehmann and
Nikolaou, (2009) concluded that the ability of banks to settle their obligations within a given time period is
called funding liquidity. Unexpected withdrawals from depositors are likely to exceed the available amount of
cash; such unbalances would cause fall in the bank liquidity i.e. asset liquidity risk and funding liquidity risk.
Hence, bank maturity transformation risk arises from the mutual interaction of the above two liquidity risk.
Maturity transformation reflects that banks cannot fulfill the unexpected withdrawals of depositors.
Objective of the Study
To identify the determinants of bank liquidity in Pakistani commercial Banks
Review of Literature
Horne and Wachowicz, (2000) said that more liquidity creation for general public can cause higher risk because
a maturity transformation risk can arise and cannot satisfy the claim of depositor’s demand. Bryant (1980),
Diamond and Dybving, (1983) were presented first model regarding banks runs, Deposit insurance, Liquidity
and explored that the main role of banks are providing liquidity. Moore, (2009) explored that it is essential for
banks to keep enough liquidity, so banks can meet the depositors claim without any barrier otherwise bank runs
may occur. If there is shortage of bank liquidity then banks borrow funds from other banks or central banks to
fulfill the depositor’s claims. If depositor’s claims are not fulfilled by the banks then depositors lose their trust on
banking system and ultimately it exposed to runs on banks. Borodo et al., (2001) states that crisis are the
inherent part of the business cycle, when the economy goes into depression, so this will impact the return of
business units and face difficulties in repaying loans and eventually, bank non-performing loan increase.
Therefore, Banks in one end will face liquidity of their assets (loans) and on the other stand it would be liquidity
of their liabilities (deposits) and banks are going to bankruptcy.
Theoretically there are two contradictory views about bank capital and liquidity. According to the first
view there are further two different approaches under which bank capital may hinder liquidity, the financial
fragility structure and the crowding-out of deposits hypothesis. The first approach refers to lower bank capital
leads higher liquidity (Diamond & Rajan, 2000, 2001), whereas higher bank capital leads crowed-out deposits
and by this means leads lower bank liquidity (Gorton & Winton, 2000).The crowding- out of deposits hypothesis
refers to shift the investor’s money and short-term deposits into bank capital. The investments on capital are not
easily converted into cash and cannot be withdrawn as desired and this will reduce bank liquidity. Under The
second view, the risk absorption hypothesis, which referred to higher capital favors to generate more liquidity
(Diamond & Dybvig, 1983) and (Allen & Gale, 2004).
Iannotta et al., (2007) explained that ‘too big to fail’ argument, under this argument large banks have no
preference on liquidity and small banks have maintain high liquidity. If large banks need liquidity then they can
easily approach to external financing within a given time period but it is impossible for small banks to access
easily external financing. In this connection, large banks have low cost of funding, because other financial
institutions and central banks have trust on large banks as compared to small banks. The central bank advances
loans for any bankruptcy fall upon by large commercial banks; therefore, large banks take benefit from an
inherent assurance and invest in riskier asset.
Kiyotaki & Moore, (2008) explained that large banks prefer low liquidity because in a situation of cash
shortage central bank give advance to them. In contrary, Rauch et al., (2009) and Berger and Bouwman (2009),
explored that smaller banks are likely to be stress on intermediation process hence, they have smaller amount of
liquidity.
According to Louzic, Vouldis and Metaxas, (2011) said that moral hazard of “too big to fail” hypothesis
states that large banks undertake excessive risk i.e. investment on risky assets and more loans to borrower. So
large banks cause higher NPL’s and eventually they are going to bankruptcy as in case Lehman Brothers.
Moreover, Keeton and Morris, (1987) have first proposed “Moral hazard” hypothesis. They said that banks
increase their loan portfolio as compared to capital investment so, their Non-performing loan will rise and finally
the large financial institutions are going to bankruptcy. Bloem and Gorter, (2001), investigated that NPL may
disturb all business units, but the most considerable influence is on financial institutions which are likely to have
large loan portfolios. More non-performing loans reflect loss of depositors and foreign investors and this leads to
liquidity problems and ultimately this may create a cause of bankruptcy. Consequently, the NPL’s has an adverse
influence on banks liquidity.
Molyneux & Thorton, (1992); Goddard et al., (2004) concluded that if banks hold high liquidity leads
high opportunity cost and ultimately low profitability for banks. Hampel et al., (1994) also support this argument.
Moreover, Myers and Rajan, (1998) said that no doubt more liquidity increase the ability of a bank to meet the
claim of depositors, this will cause rise in opportunity cost and negatively influence on bank profitability and
also raise the solvency risk for commercial banks. Owolabi et al., (2011) said that there should be trade-off
between profitability and bank liquidity. Moreover, Bordeleau and Graham, (2010) concluded that a limit should
be maintain for holding liquidity, if banks crossed this limit the profitability will be decline. They conclude that
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.1, 2017
49
banks should keep a trade-off balance between bank liquidity and ROE.
Painceira, (2010) states that liquidity preference for commercial banks differ during different business
cycles. The researcher says that during economic expansions, the investment opportunities will rise in the
country. The commercial banks have greater confidence on business units. Therefore, the liquidity preference
decrease and expand loan able funds to borrowers. While on other hand bank hold more liquidity during
economic downturn. As in Pilbeam, (2005) in line with the above theoretical relationship argued that there is rise
in loan when the economy of a specific country is on higher side. Bordo et al., (2001) said that crunches are a
common part of the business cycle. When the economy is going towards downturn then profitability of business
units will tend to decrease. Therefore, in that situation customers could not repay loans, and depositors to
perceive high solvency risk. Ultimately, the NPL’s will increase and depositors will try to withdraw their bank
deposits to protect their wealth. Therefore, a problem of maturity mismatched occurred and banks are going to
bankruptcy. Gavin & Hausmann, (1998) said that if economic condition of a country towards downturn or
crunches in business operations, this will result to reduces borrowers’ capability to meet debt. The banks NPLs
will increase and eventually banks are going to bankruptcy while on the other hand the economic prosperity in a
country tends to increase borrower’s capability to meet debt obligation and at last banks NPL will decrease.
Gil-Diaz, (1994) explains that the conventional role of a bank is accepting deposits and then undertakes
loans against the specific percentage of deposits. In an unbalanced economic situation where inflation rate is
high tends to increase in interest rate and in such situation borrowers cannot repay loans because real incomes
fall. Therefore, this may cause to fall down the economic activity of a specific country. Huybens and Smith,
(1998, 1999), explored that if inflation increase in a country which tends to decrease the returns of all business
units. In such specific situation, the banks makes less loans, resource allocation is less efficient, as well as
reduces the intermediary activities of banks. Hence, rise in inflation in a country will increase the bank liquidity.
Hypothesis Development
Based on the literature reviewed this study has the following hypothesis
Bank-specific Hypotheses
H1: Bank Capital has positive/negative and significant impact on banks liquidity.
H2: Bank size has negative/positive and significant impact on banks liquidity.
H3: NPL has a negative and significant impact on bank liquidity.
H4: Profitability has a negative and significant impact on bank liquidity.
Macro specific Hypotheses
H5: GDP has a negative and significant impact on bank liquidity.
H6: Inflation rate has a positive and significant impact on bank liquidity.
Variables of the Study
Bank liquidity measure is very important because financial institutions that fail to meet the depositors claim may
face illiquidity that result the commercial banks are going to bankruptcy. The liquidity ratio approach uses
various ratios to determine changes in liquidity. Moore, (2010), Rychtarik, (2009) and Praet& Herzberg, (2008)
have used liquidity ratios i.e. liquid assets to total assets. Researchers say that liquidity gape approach is more
confusing because there is no standard method to measure bank liquidity.
Description of Variables Table 1
Variables
Definition
CAP Shareholder’s equity to total assets
BS Natural log of total assets
NPL Non-performing loan to gross advances
ROE Earning after taxes to total equity
GDP (%) Annual growth rate of GDP
INF Consumer price index
Sample and Data Collection
A convenience sampling method was used to collect the data from all commercial banks of Pakistan. Target
samples are commercial banks that are listed with State Bank of Pakistan. Initially we have considered all listed
commercial Banks as a sample for the study. However, to make the balance panel data, we have been excluded
some of the commercial banks, because they were established in later years. The final sample of this study has
included 31 commercial banks of Pakistan for the period of 10 years from 2005 to 2014 and total observations
for this study was 310. Moreover, the data regarding macroeconomic variables is gathered from the World
Development Indicator (WDI).
Econometric Model
In order to empirically examine determinants of bank liquidity, researcher used linear multivariate regression
which has been extensively applied in the previous finance literature:
BS
NPL
ROE
GDP
INF
Where,
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.1, 2017
50
• BL = Bank liquidity
• α = Constant
• = Coefficient
• CAP = Bank capital
• BS = Bank size
• NPL = Non- performing loan
• ROE = Profitability
• GDP = Gross domestic product
• INF = consumer price index
•
Ɛi, t errorterm
• i = Commercial Banks of Pakistan
• t = Time t
Three methods are used in panel data analysis, i.e.” common effect”, “fixed effect” and “random effect model”.
For the every model, there is a distinct way to test each model as well as confirm their validity.
Empirical Findings
In this section the statistically results of the study are given. At first descriptive statistics is provided in table 11.
Afterwards the correlation matrix is provided in table 111 which shows the Pearson correlation coefficient
between the variables. After correlation matrix the Diagnostic test are given for each model as Likelihood test
and Housman Test for the best fit model for this study.
Table 2 shows the descriptive statistics for the study sample including 310 observations for each
variable. The average value of bank liquidity of commercial banks listed in SBP is 12.65% and standard
deviation is 9.6% which shows that there is low variation in bank liquidity ratio. The minimum and maximum
values of BL of Pakistan are ranged from 1.27% to 40.17%. The average value of CAP is 13.87% and ranged
from 0.2% to 47.83%. The standard deviation of CAP is 12.03% which reflects that there is low variation of
bank liquidity of Pakistan from its mean value. The average value of bank size is 18.07 million (converted into
log). The maximum and minimum values of bank size are ranged from 11.61 to 20.52 and standard deviation is
2.0
6% which shows little dispersion of bank size from its mean value. The mean value of NPL is 11.40%
and standard deviation is 8.03 which reflect that a little dispersion of NPL among Commercial Banks of Pakistan
from its mean value. The maximum and minimum values of NPL are ranged from 40.83% 5.14%. The mean
value of profitability is 13.64% and standard deviation is 8.35% which shows that a little dispersion of
profitability among banking sectors of Pakistan. The values of maximum and minimum are ranged from 1.72%
to 31.67%.
The average growth of GDP from 2005 to 2014 is 4.41% and standard deviation is 2.37% which shows
little dispersion from its mean value. The maximum and minimum values of GDP are ranged from 2.40% to
6.60%. The mean value of inflation rate is 10.76% which is more than the mean value of GDP. The standard
deviation of INF rate is 4.06% and this will reflects that there is a little dispersion of mean value of inflation rate.
The maximum and minimum values of inflation rate are ranged from 6% to 20.28 %.
Descriptive Statistics Table 2
Mean Maximum Minimum Std. Dev.
BL 0.12655 0.4017 0.0127 0.0960
CAP 0.1387 0.4783 0.0020 0.1203
BS 18.007 20.5263 11.610 2.0639
NPL 0.1140 0.4083 0.0514 0.0803
ROE 0.1364 0.3167 0.0172 0.0835
GDP 0.0441 0.0660 0.0240 0.0137
INF 0.1076 0.2028 0.0600 0.0406
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.1, 2017
51
Correlation Matrix Table 3
BL CAP BS NPL ROE GDP CPI
BL 1
CAP 0.6024 1
BS -0.4950 -0.6065 1
NPL 0.3730 0.3451 -0.2036 1
ROE -0.1350 -0.3241 0.3604 -0.1460 1
GDP 0.1582 -0.0385 -0.1526 -0.1529 0.2002 1
INF -0.0114 0.0221 0.0139 -0.0489 -0.1268 0.4634 1
The results of correlation matrix revealed that there is no existences of correlation between variables i.e. lower
than 0.80. Therefore, we can conclude that there is no problem of multicollinearity between all explanatory
variables.
Likelihood Test
This test is applied to find out which model is appropriate; common or fixed effect model. The null hypothesis
(Ho) for the test is that all the cross sections have common intercept and the alternative hypothesis is that
intercept is different for each cross section. The result is given in following table.
Likelihood Test Table 4
Effects Test Statistic d.f. Prob.
Cross-section F 18.6751 (30,273) 0.0000
Cross-section Chi-square 345.9184 30 0.0000
From the above table, the probability of cross section is significant, which means that the appropriate model is
fixed effect model as compared to random effect model.
Houseman Test
Houseman test is the most efficient way to select best model between fixed effect and Random effects.
Hauseman test Table 5
Test Summary Chi-Sq. Statistic
Chi-Sq. d.f. Prob.
Cross-section random 91.2580 6 0.0000
The result of Houseman test shows that the p-value of chi square is significant which reflects that fixed effect
model is the more efficient model than random effect model. Hence, this study is considering fixed effect model
as their final model to be analyzed and The results of fixed model as
Linear (Fixed Effect Model) Table 6
Variables Beta Standard Error t-Statistic Prob.
C 0.4260 0.0566 7.5170 0.0000
CAP 0.0850 0.0430 1.9767 0.0491*
BS -0.0192 0.0026 -7.2785 0.0000**
NPL -0.0892 0.0389 -2.2927 0.0226*
ROE 0.0296 0.0394 0.7528 0.4522
GDP 0.7029 0.2235 3.1447 0.0018**
INF 0.0901 0.0694 1.2989 0.1951
R Square 0.8246 Observations 300
A.R.Square 0.8015 (Prob) F-Stats 0.0000
(**, and * denote significance level of 1% and 5% respectively)
Discussions and Conclusions
As for CAP, we have found that a statistically significant relationship with BL at 1% of significance level. The
coefficient value is positive i.e. 0.0850 which means bank capital rises by 1%, the BL increases by 8.50%. The
finding of this study about CAP and BL are in line with the risk absorption theory provided by Diamand &
Dybving, (1983) and Allen and Gale, (2004). Moreover, result of this study regarding CAP and BL are also
relevant with the empirical findings of G.Alger and I.Alger, (1999); Chagwiza, (2014); Tseganesh, (2012);
Aymen Ben Moussa, (2015); Bunda and Desqui, (2008); Vodova, (2011b) and Cucinelli, (2013). Therefore, the
hypothesis of this study (H1) saying that “There is a positive and significant impact of CAP on BL in Pakistan”
is accepted.
Bank size has a statistically significant and negative relationship with bank liquidity at 1% significant
level. The coefficient value is i.e. -0.0192 which means that BS rises by 1%, then BL decreases by 1.92%. The
result is in line with the hypothesis “two big to fail” by Iannotta et al. (2007). Hence on the basis of this
hypothesis large banks tend to hold less liquid assets and invest in riskier assets through implicit guarantee. In
case of liquidity shortage, large banks access to Lender of the Last Resort (Central Bank) for advances to