Simultaneous System Approach
1. Introduction
Rapid expansion in bank loans to the private sector is a fairly common occurrence (Adolfo,
Dell'Ariccia, & Levchenko, 2009). In terms of the business cycle, financial accelerator processes
can interpret those episodes reasonably well, which is favorable investment prospects and robust
economic activity drives asset prices higher, increasing borrowers’ creditworthiness and allowing
them to borrow more against collaterals with higher value. As a result, credit is procyclical and
expands in synch with income. Furthermore, countries climb up the financial development ladder,
resulting in an upward trend in the Credit-to-GDP ratio. Variables that drives the financial
development may cause credit growth to surpass income growth. For example, credit booms or
periods of above-trend growth in the credit-to-GDP ratio could be fostered by financial deregulation
(consisting of the removal of capital account limits), increased competition, and financial
innovation (Igan & Pinheiro, 2011).
The latest data from the State Bank of Vietnam represents that, along with the synchronous
implementation of solutions, by the end of December 2021, credit to the economy increased by
12.97% compared to the end of 2020. In addition, credit to all the economic sectors had higher
growth rates than in the same period in 2020. In which, the fields of agriculture, rural areas,
supporting companies, and enterprises applying technology, the growth was much higher than the
general growth signal (Anh, 2022). As a result, rapid credit growth has provided numerous
advantages, for instance, assisting in the channeling of domestic and foreign savings to individuals
and investors, as well as promoting the development of the financial sector and economic growth.
While increased credit availability usually stimulates economic growth by channeling savings into
investment, rapid credit growth boosts worries about prudential hazards as well. Prudential risks,
which are the dangers to financial stability arising from bank financial positions, can manifest
themselves at both the micro and macro levels. Considering the micro-level, the rapid growth of
loan portfolios may cause capacity limits, including managing risks, acquiring information, or
assessing application quality. This issue begins to bind, resulting in new loans being created without
managing and screening risk properly (Udell & Berger, 2004). Considering the macro level, growth
can include strategic competition problems where banks engage more risks or financial institutions
are more integrated, increasing the riskiness of the system entirely. Alternatively, it may entail
dependence on the same asset categories and marginal loans, which are loans provided to riskier
borrowers who are possibly more vulnerable to shocks that are correlated among borrowers. Several
studies, for example, specify the mortgage credit bubble in the United States as one of the culprits
in some global financial crises since it increased the financial system's susceptibility to a single
shock, which is a drop in house prices (Igan, Laeven, & Dell'Ariccia, 2008). As a result, high credit
growth events can reduce loan quality, harm bank soundness and increase systemic risk.
However, the rapid expansion of credit has sparked worries about macroeconomic and prudential
hazards. Credit booms in industrial and emerging market economies have been linked to
unsustainable domestic demand expansions, asset price bubbles and overheating (Igan, 2007).
Financial sector troubles, such as loan losses amid a prolonged recession or after a significant
exchange rate depreciation, cannot be ruled out. The importance of these hazards in Vietnam and
the role of public policy in mitigating them are critical considerations for policymakers.
This proposed relationship between credit growth and bank soundness is quite magnetic. Similarly,
credit growth both influences and is influenced by bank soundness. Whereas most theoretical
models forecast a negative relationship between credit growth and bank soundness, the sign of the
feedback effect is unclear. It is possible that bank soundness influences credit growth favorably
since sounder banks have a greater volume to handle risks or position additional workers and may
expand quicker than others. Alternatively, the feedback effect could be detrimental since weaker
banks become more aggressive and hold more risks in a last-ditch effort to survive (Igan &
Pinheiro, 2011).
This report aims to propose the risks associated with rapid credit growth, with a focus on the
position of bank soundness as a predictor of credit growth in Vietnam. The econometric study is
built on a framework of simultaneous equations, with credit growth as one of the variables and bank
soundness is depicted as being interdependent with one another and with numerous macroeconomic
variables as well as bank-specific considerations. In addition, this study examines two hypotheses,
the first one is banks are weakened by rapid credit growth and the second one is whether credit
grows more quickly in sounder banks. From these two hypotheses, the significance of the prudential
risks linked with credit growth can be comprehended and specified.
Due to the lack of research on credit growth and bank soundness, the goal of this paper is to raise
awareness of the risks arising from credit growth and its association with bank soundness in
Vietnam. Apparently, there are only a few papers analyzing the credit growth and bank soundness
with each other only within the Vietnamese banking scope. Moreover, existed paper does not
answer the question of whether sound or less sound banks grow more rapidly. In contrast with other
research, this study uses a different way to calculate Distance to Default by using the Merton model
as a way to evaluate bank soundness.
The study comprises five sections beginning with the introduction illustrating the framework of
Vietnam’s credit growth and bank soundness. Secondly, the literature review represents a detailed
analysis on existed papers examining these two research questions. Section 3 presents techniques
for incorporating the aforementioned aspects into the model. Data description and specification tests
are brought out to prove the reliability of the model. Section 4 will represent the results discussion
of the findings. Finally, the conclusion puts the paper to an end with a research summary, limitation
and recommendation and policy implication is provided to suggest ways to help Vietnamese banks
avoid as many risks as possible.
2. Literature review
2.1. Rapid credit growth has a negative impact on banks
Similar to many countries around the world, Vietnam suffered from the extreme impacts of the
2008 global financial crisis, as well as the public debt crisis in Europe which began in 2010.
Furthermore, the economy’s internal turbulence has raised, with significant trade deficits, budget
deficits, growing high inflation, public debt, and weak economic development. These problems, to a
certain extent, have made the banks in Vietnam face several inconstancies, which consist of non-
performing loans and liquidity issues (Hao, 2012). The underlying reason for the deterioration of
Vietnam's banking system's soundness has been universally agreed upon: too much credit expansion
from 2007 to 2010, as well as the rise of the size and number of banking systems. This can be
explained by the reduction in marginal benefits caused by the credit boom in Vietnam’s economic
sector (Do, Nguyen, & Le, 2017). This finding of the impacts of the credit boom on the soundness
of Vietnamese Commercial Banks pointed out the features related to the effects that the credit
growth between 2007 and 2010 had on the soundness of Vietnam’s banking system, which includes
the quality of bank assets, liquidity risks, banking system profitability and capital adequacy. They
came to the conclusion that the credit growth which lasted from 2007 to 2010 had a significant
impact on macroeconomic stability generally and the soundness of the banking sector particularly.
This is a result of the banking system’s widely ranged growth model of Vietnamese banking
systems. Moreover, the system has been overly focused on increasing total assets through credit
growth, with little consideration for the quality of bank lending and bank control.
In the study by Tamirisa and Igan (2007) about the credit boom and bank soundness in Europe, they
suggested that rapid credit growth in the New Member States did not weaken banks substantially
but it started to rely less on bank soundness, which is an effect that is declared in banks and credit
markets which are expanding strongly. The finding indicated that rapid credit growth in the NMS
had not noticeably but had been connected with increasing prudential risks empathizing the
significance of proactive, risk-based supervision. That supervision can assist in keeping the dangers
of rapid credit growth manageable, at the same time maximizing the advantages of rapid credit
growth for financial and economic development. The outcomes of this research suggested that a
diverse supervisory response is a must for all nations, depending on the magnitude of the prudential
concerns related to rapid credit growth. Rapid credit growth in some countries has weakened banks
or weak banks are enhancing quickly, a powerfully built policy reaction may be required. In
addition, according to SSI, income after tax of the top 14 banks in Vietnam increased by 43% in 9
months of 2021 compared to the same period last year. As a matter of fact, the banking industry
could have done even better in the fourth quarter of 2021 because this was when the States Bank
loosened the room for credit at the highest ceiling level for many credit organizations. Thanks to the
assistance of the credit growth limit which was provided to several banks at the end of December of
2021, credit growth reached a total of 13.2%. Vice Deputy Governor of the State Bank of Vietnam
said that for the whole year, with a remarkable increase reflecting the recovery of capital needs of
businesses and the economy at the end of 2021, credit growth in 2021 could have increased up to
14% (My, 2022).
2.2. Credit grows more rapidly in sounder or less sound banks?
Recognizing the possibility of a two-way causal relationship between loan growth and bank
soundness, it is necessary to take into account macro-level research specifying the drivers of credit
growth (Cottarelli, Dell'Ariccia, & Vladkova-Hollar, 2005) and (Egert, Backe, & Zumer, 2006).
Moreover, a micro-level study points out the effect of credit growth on bank soundness (Maechler,
Miltra, & Worrell, 2010) and a paper highlighting the position of bank soundness as an element
driving credit growth (Dell'Ariccia, Detragiache, & Rajan, 2005) (Zicchino & Erlend, 2006). They
found that credit growth had a negative influence on bank soundness, but only in the early section
of the sample period which was 1995 to 2000. During that era, growth is primarily driven by
sounder banks, but this relationship weakens from 2001 to 2005. Specifically, in terms of loan
growth, less sound banks have begun to catch up with sounder banks. The deterioration of the link
between credit growth and bank soundness may point to higher risks related to rapid credit growth,
since loan maker is equally likely to occur in less sound (and potentially susceptible banks). This
finding holds up when various measures of bank soundness and model assumptions are used.
While the data are separated according to the rate of aggregate credit growth, the threshold effect
becomes apparent. The result that credit growth weakened banks from 1995 to 2000 and that credit
growth became less dependent on bank soundness from 2001 to 2005 is stronger during credit
booms than during moderate-growth years. In fact, there is evidence that weak banks may expand at
the same rate as strong banks during booms. Furthermore, banks that develop at a greater rate than
the typical bank in a particular country show a negative correlation between bank soundness and
credit growth. These could be taken as helping the idea that to survive, less sound banks issue
riskier loans. Noticeably, the assessment of credit growth to bank soundness is significantly
negative over the entire sample period only during credit booms, which supports the notion that
banks have a tendency to dip into marginal borrower pools in credit boom periods. As a result,
deterioration of loan quality occurs, this explains why credit booms frequently go with financial
crises.
The study about credit growth and bank soundness by Igan and Pinheiro (2011) used a specific data
set including banks in 90 countries between 1995 and 2005 to model credit growth and bank
soundness. In this report, bank’s credit growth is evaluated as the annual proportion change in total
outstanding loans, but their soundness is estimated by distance to default. Looking at the results,
sounder banks had an incentive to grow quicker during the period, nevertheless, compared to the
late 1990s, the rate of credit growth from 2001 to 2005 was more independent of bank soundness.
Having this result, it can be interpreted that sounder banks enhances credit quicker in the 1990s,
however, they started to get worn out by the fast speed of growth. Forward to the 2000s, the less
sound banks kept on expanding credit approximately as fast as the sound banks. They concluded
that rapid expansion from weaker banks may eventually jeopardize the stability of the banking
system in subsequent years, because weak banks tend to make effort to surpass their starting issues
by taking part in high risk/high return projects.
3. Methodology
3.1. Data collection
The data is gathered from 10 Vietnamese commercial banks over 10 years starting from 2012 and
provided with information from the annual financial reports. The chosen banks for the study should
be active until the end of 2021 as well as represent the Vietnamese banking system. In addition, the
total assets of the chosen banks have taken account for over 85% of the total assets of the whole
system in 2021. The macro statistics collected are found in the information disclosure by State Bank
of Vietnam, the International Monetary Fund and Worldbank during this period.
3.2. The Empirical Model
On a theoretical level, the relationship between credit growth and bank soundness might be either
positive or negative. Having their higher capital cushions and presumably superior risk
management, stronger banks are projected to have a competitive advantage in reaching the credit
demand. As a result, sounder banks may expand more credit. However, if loan portfolios develop
quicker than banks’ ability to assess and manage risks, credit risks may rise and loan quality
deteriorates, resulting in increased non-performing loans and less profit. One may also argue the
contrary: for the sake of surviving, weaker banks may be required to offer loans aggressively. The
hazards connected with quickly rising loan portfolios would be more prominent in that
circumstance. Overall, the relationship between credit growth and bank soundness carries on being
an econometric mystery. The two-way relationship between bank credit growth and soundness can
be investigated using a simultaneous equation model.
3.2.1. Dependent Variables
For this paper, credit growth and bank soundness are modeled as functions of each other also as
several macroeconomic and bank-specific features. Specifically, individual banks’ credit growth is
evaluated as the annual percent change in total outstanding loans, while the distance to default
measures the soundness of banks. The model includes lagged dependent variables to account for
credit growth persistence and distance to default. By successively examining the significance of
several macroeconomic and bank-specific variables found in the recent studies as structural
predictors of credit growth and bank soundness, a parsimonious baseline specification was
established.
Specifically, distance to default has gradually become a famous measurement for the soundness of
the bank (Danmarks Nationalbank, 2004). Its reputation originates from the truth that it is tied to the
possibility of default, that is, the likelihood that the value of assets falls below the value of debt.
Furthermore, this paper will use the Black Scholes model to identify each bank’s distance to
default.
According to the Black Scholes formula for a European call option, the distance to default is
estimated by:
( )(
St
)
2
σ
ln + r− T
K 2
d 2=
σ√T
with St is the value of the underlying asset, K is the strike price, r is the risk free rate, σ is the
volatility and T is the time period.
However, some factors will then be replaced (Dar, Anuradha, & Shahid, 2019). The risk free rate
will be replaced by the expected continuously compounded return on the value of the firm μV . The
value of the underlying asset a time t will become the value of the firm at time t is V. The strike
price turns into the debt value D. Finally, the volatility is replaced by the volatility of the firms’
value σ V .
The new distance to default formula is provided by:
( )( )
2
V σ
ln + μV − V T
D 2
d 2=
σV √ T
with V as value of firm at time t, D as debt value, μV as the expected continuously compounded
return on value of the firm, σ V as the volatility of the firms value and T is the time period.
3.2.2. Independent Variables
Macroeconomic factors reflecting demand-side indicators of bank credit growth and the effect of
macroeconomic states on bank soundness served as an initial step for the baseline formulation.
According to the majority of macro-level findings of credit growth, the set of variables include:
GDP per capita: demonstrating the catching-up phenomenon, in which credit growth is
slower in nations which are in a higher state of economic and institutional development.
Real GDP growth: indicating the credit capital demand of the economy. When the GDP
growth is high, investment opportunities will increase, as a result, the demand for bank
lending also escalates. This case is accurate for Vietnam’s bank-based financial system. It is
expected to be positively correlated with the demand for bank loans.
Real interest rates: specifying how much credit capital of the economy needs. If real interest
rate increases, the real borrowing cost will increase as well. Hence, real interest rate has a
tendency to be negatively correlated with the demand for bank loans from commercial
banks.
The change in real exchange rate: illustrates the requirement for credit in foreign currency
loans, simultaneously, exchange rate variation also has an effect on the safety of banks.
The macroeconomic variables mentioned are expected to reflect the risks that banks face,
consequently, might have an impact on their soundness.
Secondly, the micro variables that demonstrate the business of each Vietnamese bank are selected
to give a description of the elements that have an impact on the soundness of the banks. In terms of
the soundness of banks, these features are contemplated, consisting of profitability which is proxied
by net interest margin, banks’ size (log of total asset), liquidity (the liquidity ratio) and the
efficiency (cost to income ratio), and foreign and public ownership (the proportion of the share of
capital owned by foreign shareholders and the government, respectively). These factors are used as
explanatory variables for bank credit growth and soundness’s equation. The more profitable,
efficient and liquid, the sounder the bank and the faster the credit expansion rate (Igan, 2007).
Furthermore, it is expected that the soundness of banks and credit growth would be positively
correlated with the size and foreign ownership, meanwhile, negatively correlated with state
ownership. These variables may catch the impact of financial and several institutional reforms on
banks’ incentives and their capability to allow the private sector to borrow.
Variables Description and calculations Symbol Sources
Bank’s The rise in demand for loans. BankCreditGrowthit Annual
Credit Credit growth rate = consolidated
Dependent Variables
growth (Outstanding Loans at time t financial statement
– Outstanding Loans at time
t-1)/ Outstanding Loans at
time t-1
( )( )
Bank V
2
σ DistanceToDefaultit Annual
Soundness ln + μV − V T consolidated
D 2
d 2= financial statement
σV √ T
GDP per GDP per capita is gross GDPperCapitat [Link]
capita domestic product divided by
midyear population.
Macro Independent Variables
Real GDP The cycle of the economy, GDPgrowtht [Link]
growth annual percentage growth
rate rate of GDP at market prices
based on constant local
currency.
Real Real interest rate is the RIRt [Link]
interest lending interest rate adjusted
rate for inflation as measured by
the GDP deflator.
Exchange Annual change of the real ∆RERt [Link]
rate exchange rate
Cost to Represents the relationship CostToIncomei,t Annual
Income between the income and the consolidated
cost to achieve that income financial statement
while describes the
effectiveness of the
Characteristics of Vietnamese banks
operations.
Cost to Income = Operating
cost/Operating Income
Interest The net interest margin, InterestMargini,t Annual
Margin measures the profitability of consolidated
banks. financial statement
Net Interest Margin = (Net
Interest Income- Net Interest
Expense)/Average Interested-
Earning Assets
Liquidity Liquidity is a measure of the Liquidityi,t Annual
cash and other assets banks consolidated
have to promptly pay bills financial statement
and meet short-term business
and financial obligations.
Liquidity ratio = Liquid
assets/Total Assets
Size Size of the assets Sizei,t Annual
Ln(Total Assets) consolidated
financial statement
Public The proportion of share of Publicit Annual Report
capital owned by the
government.
Foreign The proportion of share of Foreignit, Annual Report
capital owned by the foreign
individuals or institutions.
Table 1: List of variables
3.3. Model Specification
Equation 1:
BankCreditGrowthit = f (BankCreditGrowthi,t-1, GDPperCapitat-1 , GDPgrowtht-1, RIRt-1, ∆RERt-1,
DistanceToDefaulti,t-1, CostToIncomei,t-1, InterestMargini,t-1, Liquidityi,t-1, Sizei,t-1, Foreignit, Publicit)
Equation 2:
DistanceToDefaultit = f (BankCreditGrowthi,t-1, GDPperCapitat-1 , GDPgrowtht-1, RIRt-1, ∆RERt-1,
DistanceToDefaulti,t-1, CostToIncomei,t-1, InterestMargini,t-1, Liquidityi,t-1, Sizei,t-1, Foreignit, Publicit)
With i indicates individual banks and t denotes the year index. BankCreditGrowth is the proportion
change in real bank credit to the private sector every year. The real interest rate is denoted by RIR
and ∆RER represents the percentage change in real exchange rate over the course of the year.
Furthermore, CostToIncome and InterestMargin refer to the cost-to-income ratio and the net interest
margin, respectively. Public and Foreign are indicators of public and foreign ownership.
The three-stage least squares approach can be used to estimate these two equations together.
Arellano (1990) concluded that the three-stage least squares method (3SLS) is a useful regression
model employing panel data with short time dimension and containing lags of the dependent
variables. In addition, the benefits of using 3SLS are shown in this context. 3SLS is used for a
simultaneous equation setting, which is different than the usual way for calculating single-equation
dynamic panel models proposed by Arellano and Bond (1991). Secondly, considering the cross-
equation correlation, 3SLS produces more proficient results for contemporaneous equation systems
than two-stage least squares (2SLS). Thirdly, 3SLS lets the covariance matrix of the residuals
created by the two equations to be unrestricted. Hence, the results from the regression are
trustworthy, not drifted and more desirable than that of the two-stage method consisting of lagged
dependent variables in the model (Brooks, 2008).
4. Results and Discussion
4.1. Data Description
As you can see from Table 2, the descriptive statistics of the variables in the regression model are
presented. It is noticeable that the average bank credit growth rate of 10 commercial banks in this
study is quite moderate at 17.81% and the dispersion of the values is relatively low with a standard
deviation of 10.71%. The lowest credit growth rate recorded among 10 banks is 22.18% while the
highest is 48.94%.
Mean Max Min Std. Dev. Obs
BANKCREDITGROWT
0.18 0.49 -0.22 0.11 100
H
GDPPERCAPITA 2322 2857.6 1735.1 377.42 100
GDPGROWTH 0.06 0.07 0.03 0.02 100
RIR 0.05 0.07 0.02 0.02 100
_RER 0.01 0.04 -0.01 0.01 100
DISTANCETODEFAUL
16.08 27.89 6.76 4.93 100
T
COSTTOINCOME 0.46 0.87 0.03 0.13 100
INTERESTMARGIN 0.04 0.09 0.02 0.02 100
LIQUIDITY 0.31 0.52 0.09 0.09 100
4.71E+0 1.76E+0 4.60E+0 4.17E+0
SIZE 100
8 9 7 8
FOREIGN 0.19 0.30 0.00 0.09 100
PUBLIC 0.28 0.96 0.00 0.35 100
Table 2: Descriptive Statistics of the variables
This represents the truth of the Vietnamese banking market's structure and competition. Despite the
fact that the number of commercial banks in Vietnam is quite large, the majority of credit market
shares have been held by a small group of state-owned commercial banks. As a result, small
commercial banks have frequently struggled with loan growth, even during the system's credit
boom phase, due to potential resource, capital, and client base constraints. Furthermore, looking at
the histogram, along with the skewness of 0.28, the distribution of the credit growth rate can be
considered fairly symmetrical.
Figure 1: Histogram of Bank Credit Growth variable
On average, banks’ distance to default, using the Merton model, ranges around 16.08. The
histogram indicates that the distribution of distance to default values is also quite symmetrical. In
addition, the highest value of distance to default is 27.89 compared to the lowest of 6.76.
Furthermore, the level of dispersion of the distance to default is relatively low which indicates that
there might be an adequate level of synchronization in the Vietnamese banking system’s operations.
This would help the banking system to avoid systemic risk and be well protected from unfortunate
shocks in the economy (Do, Nguyen, & Le, 2017).
Figure 2: Histogram of Distance to Default variable
The range for correlation coefficients is between -1 and 1, with 1 indicating a positive linear
correlation and -1 indicating a negative linear correlation. The correlation demonstrates just the
relationship between the variables, not their causality. Multicollinearity in the regression model can
be caused by strongly correlated variables, affecting estimation accuracy. According to Kennedy
(2008), when the absolute values of the correlation coefficients are larger than 0.80, the correlation
between variables is strong. Meanwhile, it is proposed that when coefficients exceed 0.7,
multicollinearity occurs (Anderson, Sweeney, & Williams, 1990). Looking at Appendix 1, it shows
that some variables are substantially correlated, which are (∆RER; GDP Growth) and (Public; Size).
4.2. Regression without redundant variables
After dropping all insignificant variables, credit growth is consistent with the finding that it has a
statistically insignificant impact on the bank. At the same time, these changes do not dramatically
alter the results, the lagged distance to default variable in the credit growth equation is reported to
be statistically significant and its coefficient is negative. In terms of the research questions, this can
be interpreted that credit growth does not have a noteworthy effect on the bank soundness.
Moreover, the result also indicates rapid expansion from less sound banks, in other words, smaller-
scale banks are likely to operate more in the lending sector.
BANKCREDITGROWTH = 0.854033 + 0.252286*BANKCREDITGROWTH(-1) -
0.004546*DISTANCETODEFAULT(-1) - 0.029344*LOG(SIZE(-1)) - 0.324526*FOREIGN
DISTANCETODEFAULT = 7.908862 + 0.275431*BANKCREDITGROWTH(-1) + -
0.00045*GDPPERCAPITA(-1) + 0.904227*DISTANCETODEFAULT(-1) + -
0.333544*LOG(SIZE(-1)) + 0.727449*PUBLIC
4.3. Regression with first and second differenced variables
The distance to default variable statistically has a negative effect on the bank credit growth,
concluding that less sound banks tend to extend credit as quickly as sounder banks. Not only that,
but the same result is also seen from the soundness equation, which is the growing in credit has
nothing to do with the soundness of the bank. This finding is in line with the observations of Hilbers
and others (Hilbers, 2005), who reported no substantial decrease in financial soundness
determinants based on a general review of financial soundness indicators. Finally, the empirical
regression for the fixed model is:
BANKCREDITGROWTH = -0.391953 + 0.240040*BANKCREDITGROWTH(-1) +
0.000847*GDPPERCAPITA(-1) - 3.016388*D(GDPGROWTH(-1),2) - 0.833172*RIR(-1) +
0.677194*D(_RER(-1)) - 0.005859*DISTANCETODEFAULT(-1) +
0.151237*COSTTOINCOME(-1) + 0.821245*INTERESTMARGIN(-1) +
0.161170*D(LIQUIDITY(-1),2) - 0.031688*LOG(SIZE(-1)) + 0.308316*D(FOREIGN) +
0.073403*PUBLIC - 0.144755*TREND
DISTANCETODEFAULT = 11.832258 - 1.193708*BANKCREDITGROWTH(-1) -
0.003993*GDPPERCAPITA(-1) - 14.382358*D(GDPGROWTH(-1),2) - 2.840980*RIR(-1) +
12.281985*D(_RER(-1)) + 0.911232*DISTANCETODEFAULT(-1) +
1.546140*COSTTOINCOME(-1) + 13.356490*INTERESTMARGIN(-1) +
3.107948*D(LIQUIDITY(-1),2) - 0.232820*LOG(SIZE(-1)) - 11.717874*D(FOREIGN) +
0.976732*PUBLIC + 0.298616*TREND
5.2.4. Dividing into two sub-periods
Analyzing the regression in two sub-periods, credit growth in Vietnam appeared to be a negative,
but statistically insignificant effect on bank soundness both during 2012-2017 and 2017-2021.
Therefore, the result is consistent with the two previous conclusions this study has winded up.
Moreover, the paces of credit growth in Vietnam in two sub-periods are both dependent on bank
soundness, while showing a significantly negative connection with it. This finding is also in line
with other tests in the research, concluding that less sound banks produce a more rapid expansion in
credit. These conclusions are similar with the findings by Natalia and Deniz (Igan, 2007), however,
the independence of credit growth on bank soundness was explained by the significance change in
two periods. From their study, when the DistanceToDefault variable was no longer significant in the
second period, it meant that less sound banks could grow as rapidly as healthier banks.
Empirical regression using sample from 2012 and 2017:
BANKCREDITGROWTH = 0.608062 + 0.374609*BANKCREDITGROWTH(-1) -
0.000179*GDPPERCAPITA(-1) + 23.938060*GDPGROWTH(-1) - 0.491870*RIR(-1) -
8.964545*_RER(-1) - 0.006898*DISTANCETODEFAULT(-1) + 0.204697 *COSTTOINCOME(-
1) + 1.821127*INTERESTMARGIN(-1) - 0.041254*LIQUIDITY(-1) - 0.072785*LOG(SIZE(-1)) -
0.228074*FOREIGN + 0.183351*PUBLIC
DISTANCETODEFAULT = 10.955666 - 1.013279*BANKCREDITGROWTH(-1) -
0.003532*GDPPERCAPITA(-1) + 276.616078*GDPGROWTH(-1) + 4.722649*RIR(-1) -
79.700327*_RER(-1) + 0.880291*DISTANCETODEFAULT(-1) + 0.996364*COSTTOINCOME(-
1) - 19.066469*INTERESTMARGIN(-1) - 3.007798*LIQUIDITY(-1) - 0.880753*LOG(SIZE(-1))
- 1.055743*FOREIGN + 2.007418*PUBLIC
Empirical regression using sample from 2017 to 2021:
BANKCREDITGROWTH = -1.398016 + 0.219431*BANKCREDITGROWTH(-1) +
0.000645*GDPPERCAPITA(-1) - 5.094603*GDPGROWTH(-1) + 2.464956*RIR(-1) +
32.876351*_RER(-1) - 0.009490*DISTANCETODEFAULT(-1) + 0.262396
*COSTTOINCOME(-1) + 1.028385*INTERESTMARGIN(-1) + 0.280027*LIQUIDITY(-1) -
0.017240*LOG(SIZE(-1)) - 0.321653*FOREIGN + 0.062016*PUBLIC
DISTANCETODEFAULT = 51.61432 - 0.973990*BANKCREDITGROWTH(-1) -
0.014512*GDPPERCAPITA(-1) + 72.469036*GDPGROWTH(-1) - 41.038903*RIR(-1) -
625.635234*_RER(-1) + 0.878486*DISTANCETODEFAULT(-1) +
0.969231*COSTTOINCOME(-1) + 2.548738*INTERESTMARGIN(-1) - 0.359049*LIQUIDITY(-
1) - 0.475578*LOG(SIZE(-1)) - 1.764440*FOREIGN + 0.9785407*PUBLIC
5. Conclusion
5.1. Summary of the research
Using bank-level data collection and understanding the two-way causality between loan growth and
bank soundness, this study investigated the prudential features of credit growth in Vietnam. The
sample is gathered from 10 Vietnamese commercial banks from 2012 to 2021, largely via annual
reports and trustworthy stock websites. The input contains 100 observations in total and panel
regression models are used to provide statistical proof. By doing so, this paper combines macro-
level studies that look at the determinants of credit growth and micro-level analysis that examine
the effect of credit growth on bank soundness. Moreover, this research also provides additional
examinations by analyzing the relationship between credit growth and bank soundness in two
separate periods. According to the econometric research, credit growth in Vietnam during the last
decade has represented financial deepening as well as several macroeconomic factors such as strong
economic development, real interest rates, and exchange rate. Credit growth has also been
influenced by specific bank characteristics such as profitability, soundness, efficiency, the
proportion of state ownership. Bank soundness has generally been determined by bank-specific
variables, including size, history, liquidity, and the level of ownership by foreigners as well as the
country's degree of economic and institutional development in which the bank is.
Rapid credit growth in Vietnam has not shown any sign of weakening the bank soundness so far,
however, it has been becoming less dependent on bank soundness. This is an impact that is
especially obvious in banks and credit markets that are developing very rapidly. In other words, the
results of the investigations represent a fact that less sound banks seem to grow at a faster rate.
Weaker banks which associated with rapid expansion may have small distances to default and hence
be weak in absolute terms, creating prudential issues.
5.2. Recommendations
The discovery that high credit growth in Vietnam has not severely weakened banks but has been
connected with increased prudential risks emphasizes the necessity of risk-based and ambitious
supervision. In addition, supervision can assist keeping the dangers of rapid credit expansion
manageable while maximizing the advantages of rapid credit growth for financial development and
economic growth. The findings of this research suggest that a diverse supervisory response is
required for the entire nation, depending on the magnitude of the prudential concerns associated
with rapid loan growth. For a country where rapid loan expansion has weakened banks or weak
banks are expanding rapidly, a stronger policy reaction may be needed.
To strengthen regulatory authorities’ supervisory role, the State Bank of Vietnam must increase
commercial banks’ disclosure in terms of risk management methods and policies. Moreover, the
State Bank should mandate commercial banks to conduct stress tests on a regular basis, proposing
specific solutions in bad circumstances to reverse the credit cycle and the economy. The State Bank
should also shift from a compliance-based supervision strategy to a risk-based supervision one.
Finally, in addition to applying macro-prudential supervision to commercial banks, the State Bank
of Vietnam must improve the managers of non-bank credit institutions.
For the banks, they must raise awareness of the hazards related to rapid credit growth. Not only that,
credit risk management must be addressed, with an emphasis on determining risk appetite,
enhancing the credit risk management model, raising credit risk awareness, and measuring it. Last
but not least, commercial banks should establish a proper loan portfolio management strategy to
prevent credit growth based on highly risky industries for example, real estate and stock trading.
5.3. Policy implications
According to the econometric findings, rapid credit growth in Vietnam has not been having a
substantially detrimental effect on bank soundness thus far, however, it has grown pervasive in
recent years, with both relatively weak and sound banks’ lending at a similar quick pace (Hà, 2022).
A considerable accumulation of financial risks in the weaker banks threatens to undermine their
soundness over time. Specifically, if the weaker banks are not even systemic, their distress may be
contagious within the financial system and be associated with systemic effects (Igan, 2007). As a
result, the economic findings lead to three major policy goals in Vietnam’s banking system. Firstly,
close monitoring of fast developing banks to be certain that they have suitable risk management
systems available as well as effectively pricing risks, in line with the reinforcing of bank regulatory
instructions. Secondly, enhancements to market infrastructure and institutions that promote
sustainable credit expansion, such as credit bureaus and customer transparency of financial risk
information. Finally, the managers should have a better grasp of the inference of credit expansion
for financial stability in the framework of financial sector surveillance and macro prudential risk
assessments.
Detailed steps that can be made to improve financial sector surveillance consist of the following:
Financial soundness indicators: The findings can help enhance Viet Nam's institutional and
statistical capacities to regularly collect, analyze, and distribute internationally comparable
financial soundness indicators, they are also used to assess an economy's financial health in
a scientific, systematic, and comprehensive manner, based on five areas of interest: capital
adequacy, asset quality, profits, liquidity, and market risk sensitivity. They are
methodological techniques that aid in quantifying and categorizing the stability and
vulnerability of financial systems (Asian Development Bank, 2015).
Early Warning Systems: Early warning systems based on CAMEL (capital, asset quality,
management, earnings, and liquidity) indicators can be improved further by incorporating
indications that aid in the prediction of bank failures (Tamirisa & Čihák, 2005). In Vietnam,
the model of early warning of risks allows Military Commercial Joint Stock Bank (MBB) to
identify customers with signs of declining financial and prompt payment ability so that
banks can provide appropriate behavioral solutions to support customers handling debt
recovery. The system can early identify 40% of the problem loans before 6 months and
support MBB in good credit quality management (Minh, 2021).
Loan data source: collecting data from a wide range of source
Balance sheet analysis: provide information about the balance-sheet risks associated with
rapid loan growth, as well as their inter-sectoral dispersion of risks.
The application of the Basel II capital framework in the Vietnamese banking system should assist
increase risk-based supervision, nevertheless, the advantages of Basel II are expected to come
relatively gradually, and the short-term impact is unclear.
The implementation of Basel II will increase emphasis on risk management techniques, disclosure,
and market discipline. Basel II adoption should bring prudential laws more in line with excellent
risk management and motivate banks to create risk management systems. Most of the discussion
about Basel II has focused on Pillar 1, which is capital adequacy, the government may profit more
in the long run from implementing Pillar 2, the supervisory practices, and 3, the growing market
discipline and disclosure (Caruana, 2004).
However, the importance of Basel II in reducing macro prudential concerns should not be
exaggerated. In particular banks, the effect is heavily dependent on how the framework is
implemented. Basel II provides banks with a number of options, which, while providing the new
framework with the required flexibility, at the same time creating obstacles for micro prudential and
macro prudential supervision. The structure also provides supervisors more flexibility, for instance,
they can adjust capital requirements among banks based on the underlying risks under Pillar 2. So
that in the short term, before any further implementation, the new regulatory framework is at least
gained, extra prudential measures may be required, as discussed in the main text of this article. A
variety of factors must be considered in this regard:
Basel II is, by definition, a micro prudential structure that does not directly address macro
prudential or macroeconomic concerns. As a result, banks' risk management systems might
be unable to account for the further-round effects of their activities on other market
participants (Čihák, 2006).
Basel II's effects on credit expansion are still unknown. Banks may expand their holdings
of low-risk assets and lower capital charges, while decreasing their assets possession that
produce a greater capital charge under Basel II and place ascending pressure on lending
rates. The variables could move financing away from higher-risk sectors (such as
commercial real estate) and toward less risky sectors like residential housing. So far,
quantitative impact analyses indicate that, Basel II implementation will most likely result in
reduced in the short term, rather than greater, capital requirements. Interpretation along
with caution must be included in these findings, but they do prove in the short term that the
potential for credit growth is likely to rise rather than sink (Čihák, 2006).
It is difficult to create an effective prudential policy replying to high loan growth. Excessively
stringent restrictions would unfairly penalize fast-expanding banks that are adequately managing
risks, and could slow down Vietnam's financial depth and economic convergence. While adding to
the distortions, policy that is excessively intrusive actions may be unsuccessful, as banks may shift
business to the sector beside banking or abroad. With the inadequate significant evidence that rapid
credit growth has damaged banks, the preceding factors suggest that risk-based supervision should
be prioritized as a way of alleviating the risks associated with rapid credit development.
References
Adolfo, B., Dell'Ariccia, G., & Levchenko, A. (2009). Credi Booms: the Good, the Bad, the Ugly.
Washington: International Monetary Fund.
Anderson, D., Sweeney, D., & Williams, T. (1990). An Introduction to Management Science: A
Quantitative. Lincoln.
Anh, H. (2022). Tín dụng thúc đẩy phục hồi tăng trưởng kinh tế. Nhân Dân.
Arellano, M. (1990). Testing for Autocorrelation in Dynamic Random Effects Models. Review of
Economic Studies, 127-34.
Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence
and an application to employment equations. The review of economic studies, 277-297.
Asian Development Bank. (2015). Financial Soundness Indicators for Financial Sector Stability in
Viet Nam. Manila.
Brooks, C. (2008). Introductory Econometrics for Finance.
Caruana, J. (2004, September 23). Making diligent preparations for Basel II. From Bank for
International Settlements.
Čihák, M. (2006). How Do Central Banks Write on Financial Stability? IMF Working Paper.
Cottarelli, C., Dell'Ariccia, G., & Vladkova-Hollar, I. (2005). Early birds, late risers, and sleeping
beauties: Bank credit growth to the private sector in Central and Eastern Europe and in the
Balkans. Journal of Banking & Finance, 83-104.
Danmarks Nationalbank. (2004). Market-BAsed Risk Measures for Banks. Financial Stabiity 2004.
Dar, A. A., Anuradha, N., & Shahid, Q. (2019). Estimating probabilities of default of different firms
and the statistical tests. Journal of Global Entrepreneurship Research.
Dell'Ariccia, G., Detragiache, E., & Rajan, R. (2005). The real effect of banking crises. IMF
Working Paper .
Do, K. T., Nguyen, M. T., & Le, T. H. (2017). Effects of the Credit Boom on the Soundness of
Vietnamese Commercial. International Journal of Financial Research.
Egert, B., Backe, P., & Zumer, T. (2006). Credit growth in central and eastern Europe: new (over)
shooting stars? ECB Working Paper, 687.
Hà, T. (2022). Nhiều ngân hàng bứt phá mạnh về lợi nhuận và tăng trưởng tín dụng. Vietnam Plus.
Hao, Q. M. (2012). Những điểm yếu của hệ thống ngân hàng Việt Nam hiện nay. VNU Journal of
Science: Economics and Business.
Hilbers, P. (2005). Assessing and managing rapid credit growth and the role of supervisory and
prudential policies. IMF Working Paper.
Igan, D., & Pinheiro, M. (2011). Credit Growth and Bank Soundness: Fast and Furious?
Igan, D., Laeven, L., & Dell'Ariccia, G. (2008). Credit Booms and Lending Standards: Evidence
from the Subprime Mortgage Market. IMF Working Paper No. 08/106 (Washington:
International Monetary Fund).
Igan, N. T. (2007). Credit Growth and Bank Soundness in Emerging Europe.
Kennedy, P. (2008). A Guide to Econometric (6th ed.). Blackwell Publishing.
Maechler, A., Miltra, S., & Worrell, D. (2010). Decomposing financial risks and vulnerabilities in
emerging Europe. IMF Staff Papers, 25-60.
Minh, A. (2021). Triển khai tốt Basel 2 giúp hệ thống ngân hàng phát triển lành mạnh. Báo Chính
Phủ.
My, L. (2022). Năm 2022, tăng trưởng tín dụng còn kéo tăng trưởng lợi nhuận ngân hàng? Tạp chí
tài chính.
Tamirisa, N., & Čihák, M. (2005). Household Credit in the Czech Republic: Macro and
Microprudential Analysis, Cross-Country Experiences, and Implications for. International
Monetary Fund.
Udell, G. F., & Berger, A. N. (2004). The institutional memory hypothesis and the procyclicality of
bank lending behavior. Journal of Financial Intermediation 13(4), 458-495.
Zicchino, L., & Erlend, N. (2006). Bank Weakness, Loan Supply and Monetary Policy. IMF
Seminar Series .
APPENDIX 1: Correlation matrix of the variables
DISTANCE
BANKCREDITG GDPPERCAPI GDPGRO TODEFAUL COSTTOINCO INTERESTMA
ROWTH TA WTH RIR _RER T ME RGIN LIQUIDITY SIZE FOREIGN PUBLIC
BANKCREDI
TGROWTH 1.000000
GDPPERCAP
ITA 0.011196 1.000000
GDPGROWT
H 0.179225 -0.349608 1.000000
RIR 0.318964 0.147634 0.016089 1.000000
_RER 0.321975 -0.469273 0.762229 0.347574 1.000000
DISTANCET
ODEFAULT -0.061393 -0.532973 0.273643 -0.122979 0.306101 1.000000
COSTTOINC
OME -0.115431 -0.427380 0.233612 -0.000284 0.280840 0.427250 1.000000
INTERESTM
ARGIN 0.162073 0.103598 -0.102943 -0.082542 -0.121880 -0.414998 -0.431374 1.000000
LIQUIDITY 0.081783 -0.246899 0.087698 -0.219958 -0.020717 -0.033552 -0.069534 0.042880 1.000000
SIZE -0.126059 0.433123 -0.194629 0.032975 -0.230790 -0.112369 -0.444463 -0.228179 -0.352813 1.000000
FOREIGN -0.282431 0.275197 -0.083104 0.008277 -0.129733 -0.395242 0.002653 -0.100590 0.025573 -0.004020 1.000000
PUBLIC -0.099677 -0.037887 0.013023 -0.012573 0.017142 0.313701 -0.389222 -0.334467 -0.110422 0.762235 -0.175373 1.000000