Description
This study aims to investigate the inter-relatedness and the dynamics of banking stability
measures and offers answers for some of the related issues such as does financial stability require
the soundness of banking institutions, the stability of markets, the absence of turbulence and low
volatility? and to what extent the soundness of banking sector in the case of emerging economies can
help financial system stability.
Journal of Financial Economic Policy
Testing the interrelatedness of banking stability measures
Vighneswara Swamy
Article information:
To cite this document:
Vighneswara Swamy , (2014),"Testing the interrelatedness of banking stability measures", J ournal of
Financial Economic Policy, Vol. 6 Iss 1 pp. 25 - 45
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-01-2013-0002
Downloaded on: 24 January 2016, At: 21:48 (PT)
References: this document contains references to 56 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 340 times since 2014*
Users who downloaded this article also downloaded:
Faten Ben Bouheni, (2014),"Banking regulation and supervision: can it enhance stability in Europe?",
J ournal of Financial Economic Policy, Vol. 6 Iss 3 pp. 244-269http://dx.doi.org/10.1108/J FEP-11-2013-0059
Eric Osei-Assibey, Baimba Augustine Bockarie, (2013),"Bank risks, capital and loan supply: evidence from
Sierra Leone", J ournal of Financial Economic Policy, Vol. 5 Iss 3 pp. 256-271http://dx.doi.org/10.1108/
J FEP-09-2012-0041
J ill M. Hendrickson, Mark W. Nichols, Daniel R. Fairchild, (2014),"Bank branch location and stability
during distress", J ournal of Financial Economic Policy, Vol. 6 Iss 2 pp. 133-151http://dx.doi.org/10.1108/
J FEP-07-2013-0026
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Testing the interrelatedness
of banking stability measures
Vighneswara Swamy
Department of Finance, IBS-Hyderabad, Hyderabad, India
Abstract
Purpose – This study aims to investigate the inter-relatedness and the dynamics of banking stability
measures and offers answers for some of the related issues such as does ?nancial stability require
the soundness of banking institutions, the stability of markets, the absence of turbulence and low
volatility? and to what extent the soundness of banking sector in the case of emerging economies can
help ?nancial system stability.
Design/methodology/approach – This study investigates banking stability by structuring a
recursive micro panel vector auto regressive (VAR) model and corroborates the signi?cance of the
interrelatedness of the bank-speci?c variables such as liquidity, asset quality, capital adequacy and
pro?tability by employing a robust panel data drawn from 56 leading banks for a period of 12 years.
Findings – A signi?cant contribution of this study is in establishing that liquidity in the
banking-dominated ?nancial system is reciprocally related with asset quality, capital adequacy, and
pro?tability of the banking system and in effectively forecasting banking stability employing micro
panel recursive VAR model.
Research limitations/implications – The study could be further broadened by employing a
macro and structural VAR modelling to forecast banking stability.
Practical implications – This paper is one among the evolving body of literature that underscores
the signi?cant relationship between banking system resilience and ?nancial stability in the context of
emerging economies dominated with banking systems. Further, the forecast model is able to capture
the dynamics of banking stability with greater and appreciable accuracy.
Originality/value – The uniqueness of the study is in modelling banking stability measures in the
context of banking-dominated emerging economy ?nancial systems by employing micro panel recursive
VAR model by deriving data from 58 leading banks for the period of 12 years from 1996 to 2009 and
in offering insights in understanding ?nancial stability with comprehensive literature review.
Keywords Indicator, Crisis, Financial stability, Banks and ?nancial institutions, Banking stability
Paper type Research paper
1. Introduction
A threat to ?nancial stability anywhere in the world is potentially a threat to ?nancial
stability everywhere. As ?nancial stability and macroeconomic stability are intricately
related, ?nancial stability can be vulnerable even if there is price stability and
macroeconomic stability and hence cannot be taken for granted. However, empirical
?ndings suggest that greater market power in the banking market results in higher
instability. Although banks are better capitalized in less competitive markets, their
default risk remains higher. Particularly in the context of banking dominated emerging
market ?nancial systems where banks dominate more than 70-80 percent of the ?nancial
system, banking stability assumes greater prominence in ensuring ?nancial stability.
This study seeks to test the interrelatedness of banking stability measures in the
context of the dynamics of banking stability and offers understanding about the
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JEL classi?cation – E44, E58, G1, G21, G28
Journal of Financial Economic Policy
Vol. 6 No. 1, 2014
pp. 25-45
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-01-2013-0002
Banking stability
measures
25
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
signi?cance of liquidity of the banking sector in ensuring banking stability in the case
of bank dominated ?nancial systems particularly in the case of emerging economies.
This paper is one among the evolving body of literature that underscores the
signi?cance of banking stability towards attaining ?nancial stability. This study
analyses the concept and de?nition of ?nancial stability and in that backdrop analyses
the banking stability in the case of emerging market economy like that of India. This
paper begins by presenting in Section 2, the theoretical framework signifying the
recent approaches on de?ning and analysing ?nancial stability placing the banking
system at the epicentre of analysis. We emphasise more speci?cally how banking
stability can augment ?nancial stability in emerging market economies like that of
India that successfully came out of the recent global ?nancial crisis. We present the
methodology involving the data and its sources and research design explaining the
empirical framework and estimation of the micro vector auto regressive (VAR) model
of banking stability in Section 3. The results of the analyses with discussion on the
?ndings are enunciated in Section 4, and the conclusion and policy implications are
offered in Section 5.
2. Banking stability – theoretical framework
Notwithstanding its extensive use, ?nancial stability is dif?cult to de?ne let alone
measure. A sound understanding of ?nancial stability necessitates a conceptual
framework (Houben et al., 2004). In understanding ?nancial stability, the ?rst limitation
has been yet there is no widely accepted model or analytical framework for assessing the
?nancial stability as this it is still in its infant stage of development and practice,
as compared with, for example, the analysis of monetary and/or macroeconomic stability
(Schinasi, 2004). The concept of ?nancial stability is nebulous with no commonly
accepted de?nition. However, there have been some attempts to de?ne ?nancial stability.
Houben et al. (2004) considering ?nancial stability as a continuum changeable overtime
and consistent with multiple combinations of its constituent elements, de?ne it as the
ability to help the economic system allocate resources, manage risks and absorb shocks.
The best approach according to Allen and Wood (2006) is to de?ne the characteristics of
an episode of ?nancial instability ?rst and then de?ne ?nancial stability as a state of
affairs in which episodes of instability are unlikely to occur. Davis (2003) identi?es three
generic types of ?nancial instability. The ?rst is centered on “bank failures”, typically
following loan or trading losses, the second involves extreme “market price volatility”
after a shift in expectations and the third being the one that is linked to the second,
involves protracted collapses of market liquidity and issuance.
Borio (2003) and others take a macroprudential viewpoint and state ?nancial
stability in terms of limiting risks of signi?cant real output losses associated with
episodes of ?nancial system-wide distress. Suggesting an information-based de?nition,
Mishkin (1994) states that ?nancial instability occurs when shocks to the ?nancial
system interfere with information ?ows so that the ?nancial system can no longer do
its job of channeling funds to those with productive investment opportunities. On the
other hand, Crockett (1997) proposes that ?nancial stability refers to the stability of key
institutions and markets that go to make up the ?nancial system. Further, Issing (2003)
and Foot (2003) have suggested that ?nancial stability is associated with ?nancial
market bubbles, or more generally, with volatility in ?nancial market proxies as these
bubbles impair ?nancial markets ef?ciency; however, in and of themselves, they do not
JFEP
6,1
26
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
constitute a de?ning characteristic of ?nancial fragility, and more generally ?nancial
instability. Suggesting institutionally oriented de?nitions, Haldane et al. (2004), among
others, have proposed that ?nancial instability could be de?ned as any deviation
from the optimal saving-investment plan of the economy that is due to ?nancial
imperfections in the ?nancial sector. Similarly, Goodhart et al. (2004, 2005, 2006a, b)
and Tsomocos (2003a, b) offer de?nitions for ?nancial stability that hinges upon the
welfare effects on the economy and distributional consequences arising during periods
of ?nancial instability.
The distinctiveness of the banking stability from the perspective of ?nancial
stability is a well-recognised idea in the literature. Banking assumes special status as
they are regarded as more vulnerable to instability than other sectors. Hellwig (1991)
takes a different view on the problem and states that banks are special because they
represent the availability of funds and governments want to exercise tighter control
over them. It also argued that banks assume signi?cance in view of their tendency
towards maturity mismatches between assets and liabilities that can expose them to
possibility of runs (Diamond and Dybvig, 1983; Diamond, 1984; Diamond and Rajan,
2001). Moreover, referring to the recent times, banking is involved in heavy interbank
lending and payment markets. These interlinkages, in the absence of safety net
provisions pose the threat of propagation of risk from one bank to the other, creating
interbank contagion risk – one form of systemic risk (De Bandt and Hartmann, 2002).
While the vulnerability to runs due to problems in the interbank market represents one
source of concerns about bank instability originating from the liability side of the
balance sheet, a second source of instability relates to bank risk-taking on the asset
side. Because of their substantial ?nancing from many small, relatively uninformed
depositors and an often-existing public safety net in response to the previously
mentioned vulnerability, banks can be prone to taking on “excessive” risk. Thus, the
speciality of banking system stability is a widely recognised idea in the ?nancial
stability literature (Goodhart, 1987; Goodhart et al., 1998). Thus, to sum up, banks are
closely linked with our everyday lives and activities. Any large disruption of banking
operations will affect society as a whole. Banking stability is, therefore, crucial to
minimising the extensive economic and social impact that may arise from problems in
the industry.
The issue of ?nancial stability is quite organically linked with banking stability.
Banking stability is a yardstick to determine whether an economy is adequately strong
enough to withstand both the internal and external shocks. On the other hand, ?nancial
stability is a by-product of stability conditions prevailing in banking system, ?nancial
markets, and the real economy and amongst them, banking stability appears to be a
vital ingredient to ?nancial stability. Banking stability in itself depends on the
ef?cacies of the several parameters of individual banks, e.g. asset quality, liquidity,
capital adequacy, and pro?tability, etc. Since, banking stability gets affected positively
or negatively with the prevailing conditions in the ?nancial market and the real
economy; ultimately, it determines as to what extent ?nancial stability is ensured in the
economy by its ability to absorb the shocks. As such, banking stability can be treated
as a forerunner of ?nancial stability in an economy. Accordingly, this paper takes into
consideration a constructive viewpoint and de?nes banking stability as a state of
affairs in which the ?nancial system can; achieve ef?cient allocation of resources;
assess and manage ?nancial risks; absorb the emerging shocks; ensure smooth
Banking stability
measures
27
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
payments and remittances; enhance equilibrium by managing asset and price
volatility; and lead the economy towards bene?ts of economic welfare.
2.1 Financial stability and banking sector
Banking sector is regarded as an important sector for the stability of ?nancial systems
as the banks play a major role in money creation, investment for economic growth,
?nance for businesses and households and in payment systems. Hence, faltering
banking systems are found to be associated with hyperin?ations and depressions in
the economic history. Global ?nancial crisis has drawn the attention of the policy
makers’ of advanced and emerging economies towards banking stability and placed it
on the top of their agenda. While some analysts view banking stability to be related in
part to size and ownership structures, some others observe point to the failure of
private banks as evidence of the fragility of short-term and pro?t-oriented banking.
A stable macroeconomic environment is essential for banking sector stability, mainly
because uncertainty about macroeconomic policies and wavering fundamentals, such
as economic growth and in?ation, renders it challenging for banks to assess credit
risks accurately. Subdued economic growth, due to macroeconomic uncertainty or for
other reasons, may impair bank soundness as it reduces the debt servicing capacity of
?rms and households. There is also a signi?cant debate on the implications of
competition for banking sector stability (Anzoategui et al., 2012) and new evidence
suggests that lack of competition can lead to fragility, especially if certain banks
become too big to fail. In an econometric attempt to deal with the issue of ?nancial
system stability in a framework of duration models, using macroeconomic data, Aka
(2006) observe that banking crises are likely to occur every ten years or so, even if the
?nancial system has been stable until the year before the crises. Indeed, the risk of
banking crises builds up during a calm period. The results also reveal that factors such
as fundamentals weakness, structural characteristics, contagion phenomenon and past
experiences with banking crises play an important role.
Particularly in emerging economies, at present, the banking sector is by far the most
important part of the ?nancial system in all and is, therefore, also the main source of
risk for ?nancial stability. This is all the more so because the lack of well-functioning
equity markets confronts banks with relatively high credit risks, as bank credit is
necessary (to some extent) to substitute for equity. One of the important sources of
vulnerability that can affect ?nancial stability and lead to a ?nancial crisis can be the
weakness (such as a high level of short-term debt) in the ?nancial structure of
the economy, i.e. the composition and the size of the assets and liabilities on the balance
sheet. A ?nancial crisis follows when the demand for ?nancial assets of one or more
sectors plummets and consequently the banking system fails to meet the out?ows or
may be unable to attract new ?nancing or roll over existing short-term liabilities.
In this direction, ?nancial soundness matters much during the ?nancial crisis because
it gives some indication of how likely it is that ?nancial problems would be transmitted
into the real economy.
2.2 Banking stability measures
Measurement of banking stability has some challenges as it faces complication in
assessing, in contrast to other elements of the ?nancial system such as; securities
values, interbank relationships that can be at the origin of bank contagion phenomena
JFEP
6,1
28
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
or the values of and correlations between loan portfolios are mostly hard to measure
and monitor. As observed by Blavarg and Nimander (2002), even central banks and
supervisory authorities usually do not have continuous information about interbank
exposures. Financial stability literature provides a number of quantitative measures in
order to assess ?nancial stability. IMF (2006) recommends 12 core ?nancial sector
indicators (FSIs) for deposit takers and are categorised as capital-based FSIs,
asset-based FSIs, and income and expense FSIs. Similarly, such indicators are used as
monitoring variables used in the studies by Hawkins and Klau (2000), Nelson and Perli
(2005) and Gray et al. (2007) which focus on market pressures, external vulnerability
and banking system vulnerability. Some authors have focused on three indicators of
?nancial stability, namely, the z-score as measure of distance from insolvency, the
NPL-score as indicator of lending risk, and the probability of distress score (PD-score)
as measure of actual insolvency risk (Beck et al., 2009).
The choice of employing bank equity prices for measuring banking stability is
observed to be motivated by Merton’s (1974) option-theoretic framework toward
default. This approach has become the cornerstone of a large body of approaches
towards quantifying credit risk and modelling credit rating migrations including
Morgan’s (1999) CreditMetrics. However, ?nancial stability literature ?nds a number
of approaches that do not rely on market indicators as banking stability measures.
While Saunders and Wilson (1996) and Calomiris and Mason (1997) study deposit
withdrawals as a measure of banking stability in their study of failing and non-failing
banks during the great depression, Hasan and Dwyer (1994) measure autocorrelation of
bank failures after controlling for macroeconomic fundamentals during various
episodes of US banking history. Further, Calomiris and Mason (2000) evaluate the
survival time of banks during the great depression, with explanatory variables
including national and regional macroeconomic variables, dummies for well-known
panics and the level of deposits in the same county (contagion effect).
Bank size is considered a measure of bank stability, and its relationship has been
ambiguous in the academic literature (Boyd and Runkle, 1993; Calomiris and Mason,
2000) while de Nicolo (2000) observes a negative and signi?cant relationship.
Moreover, a higher amount of risky assets is expected to be associated with higher
bank failure and too rapid expansion of loan portfolio can lean to greater risk taking by
banks (Pain, 2003; Davis, 1993; Borio and Lowe, 2002). Literature also records market
concentration as a measure of bank stability as Beck et al. (2005) observe that market
concentration is associated with more ?nancial stability. In contrast, Boyd and
de Nicolo (2005) suggest that systems that are more competitive are conducive to bank
stability. On the other hand, Hoggarth et al. (1998) ?nd that a less competitive German
banking system seems to be more stable in their comparison of German and UK banks.
Most of the central banks across the globe have been publishing their ?nancial
stability reports (FSRs) extensively focusing on various market segments and banking
related variables. Banking ratios are widely analysed in most of such reports although
there are some differences. Some reports seem to concentrate on the banks’ performance
and risks in considerable detail, while the others take account insurance, hedge funds
and other forms of nonbank ?nancial intermediation. However, where banking is the
main form of intermediation, the available information largely depends on the level of
supervisory input in preparing the report. Some central banks, compute a banking
stability index using weighted average of sub-indicators of banking sector soundness
Banking stability
measures
29
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
including capital adequacy, pro?tability, balance sheet liquidity, asset quality, credit and
currency risk. On the other hand, some central banks calculate a ?nancial strength index
as a weighted average of partial indicators of the ?nancial soundness of banks. The
index combines six areas of ?nancial soundness indicators, namely capital adequacy,
pro?tability, liquidity, asset quality, interest rate risk and exchange rate risk.
Developing a banking stability index, IMF considers the banking system as a
portfolio, and based on the market-based probability of default of each individual bank
it estimates the joint default probabilities, i.e. the expected number of bank defaults in
the system, given that at least one bank defaults (IMF, 2008). In the Indian context,
Reserve Bank of India (RBI) employs a banking stability map and banking stability
index to assess inherent risk factors that impact the stability of the banking sector. The
indicator is based on ?ve indices, which represent the ?ve dimensions namely,
soundness, asset-quality, pro?tability, liquidity, and ef?ciency. A composite measure
of each dimension is calculated as a weighted average of a set of standardized ratios
that are relevant in assessing the dimension. Based on the individual composite indices
for each dimension, the banking stability indicator (BSI) is constructed as a simple
average of the above ?ve composite sub-indices. The higher value of the indicator
would suggest lower stability (Mishra et al., 2013). Gadanecz and Jayaram (2009),
summarizing the FSIs state that the ?nancial sector is characterised by monetary
aggregates, real interest rates, risk measures for the banking sector, banks’ capital and
liquidity ratios, the quality of their loan book, standalone credit ratings and the
concentration/systemic focus of their lending activities. All these proxies can be
re?ective of problems in the banking or ?nancial sector and, if a crisis occurs, they can
gauge the cost of such a crisis to the real economy.
2.3 Financial soundness in Indian banking system
Banking sector is by far the most central part of the ?nancial system in most of the
emerging economies and is, therefore, also the main source of risk for ?nancial
stability. Undoubtedly, ?nancial soundness of banks has a signi?cant sway on the
stability of the ?nancial system as a whole as the banking system constitutes more
than 75 percent of the ?nancial markets in India. The Indian banking system endured
the onslaught of the global ?nancial crisis and a factor that bolstered the normal
functioning of the banking system even in the face of one of the largest global ?nancial
crisis was its robust capital adequacy. Further, the core banking sector indicators for
India like; capital adequacy ratio (CAR) – Tier-1, gross non-performing assets
(GNPAs) to total loans, net non-performing assets (NNPAs) to total loans and return on
equity (ROE) have experienced downward pressure during the recent recession period
(Figure 1). On the contrary, liquid assets to total assets ratio has moved upwards
indicating the tendency of the banks to hold cash during the times of recession instead
of investing in loans or investment products.
Under Basel II, capital to risk-weighted assets ratio (CRAR) of Indian banks as at
end-March 2009 was at 14 percent, far above the stipulated level of 9 percent (Figure 2).
This suggests that Indian banks have successfully managed to meet the increased
capital requirement under the amended framework.
Furthermore, between March 2009 and 2010, there was a surge by about
0.5 percentage point in the CRAR re?ecting further strengthening of their capital
adequacy under the new framework.
JFEP
6,1
30
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3. Methodology and data
In view of the fact that assessment of ?nancial stability in general is made on a
broad-spectrum of risk factors, one cannot expect a single model to capture
satisfactorily all the risk factors originating and developing inside and outside the
?nancial system, respectively. Instead, a suite of models may be required. However, the
objective of the ensuing segment of this paper is to analyse the salient parameters of
banking sector performance and behaviour and to establish the signi?cance of banking
sector stability in the context of bank-dominated ?nancial system of Indian economy
Figure 2.
Capital to risk-weighted
assets ratio-bank
group-wise
(as at end-March)
Note:
a
Includes IDBI Bank Ltd
0
2
4
6
8
10
12
14
16
18
20
Basel-1 2009 Basel-I 2010 Basel-II 2009 Basel-II 2010
Public sector banks Nationalised banks
a
SBI group Private sector banks
Old private sector banks
Source: Report on Trend and Progress of Banking in India
2009-2010 of RBI
Figure 1.
Core banking sector
indicators for India
Source: International Financial Statistics (IFS) of IMF
0
5
10
15
20
A:2008 Q:2009:1
a
Q:2009:3
a
Q:2010:1
a
Q:2010:3
a
I
n
P
e
r
c
e
n
t
Note:
a
Quarterly figures : A - annual Q - quarter
Car Car-Tier-1
Net NPAs to Capital Gross NPAs to Total Loans
Return on Equity
Banking stability
measures
31
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
by employing a micro model of VAR. The proposed micro VAR model involves the
most relevant parameters of banking stability namely, liquidity, asset quality, capital
adequacy and pro?tability.
Data and variables
The data for the analysis is sourced from the robust database of RBI. The variables
might be considerably adopted to measure the degree of volatility and soundness in
the banking sector and thereby in?uence the ?nancial stability is listed in Table I.
For this purpose, a panel data (annual) involving weightages for the variables of bank
performance and behaviour for the period from 1996 to 2009 covering 56 commercial
banks in India is constructed.
The model
Vector auto regressions (VARs) are powerful tools for describing data and for
generating reliable multivariate benchmark forecasts. Sims (1980) advocated VAR
models as providing a theory-free method to estimate economic relationships, thus
being an alternative to the “incredible identi?cation restrictions” in structural models.
Used wisely and based on economic reasoning and institutional detail, VARs both
can ?t the data and, at their best, can provide sensible estimates of some causal
connections. Although VARs have limitations when it comes to structural inference
and policy analysis, so do the alternatives. A recursive VAR constructs the error terms
in the each regression equation to be uncorrelated with the error in the preceding
equations. This is done by judiciously including some contemporaneous values as
regressors.
Let Y
it
be an m £ 1 vector of random variables for the ith cross-sectional unit at time
t, and suppose that the Y
it
’s are generated by the following panel vector autoregressive
model of order one (PVAR):
Y
it
¼ FY
i;t21
þ1
it
ð1Þ
Y
it
¼ ðI
m
2FÞm
i
þFY
i;t21
þ1
it
ð2Þ
Variable Description
Capital adequacy
ratio (CAR)
De?ned as the amount of regulatory capital to be maintained by a bank to
account for various risks inbuilt in the banking system
Capital Adequacy Ratio ¼
Total Capital ðTier I Capital þTier II CapitalÞ
Market Risk ðRWAÞ þCredit Risk ðRWAÞ þOperation Risk ðRWAÞ
RWA – risk weighted assets
The higher the ratio the better is for the bank’s stability
Return on assets
(ROA)
Return on Assets ¼ Net Pro?t/Assets ¼ (Net Pro?t/Total Income)
*
(Total
Income/Assets)
The higher the ratio the better is for the bank’s stability
Net non-performing
assets (NNPA) to
net advances (NA)
Net NPA to Net Advances ¼ Net NPA=Net Advances
The lower the ratio the better is for the bank’s stability
Management of non-performing assets is a key to the stability and continued
viability of the banking sector
Liquidity coverage
ratio (LCR)
Liquidity Coverage Ratio ¼
ðCash þSLRþother short term investmentsÞ
Short term liabilities
The lower the ratio indicates less liquidity
Table I.
Description of key
variables
JFEP
6,1
32
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
for i ¼ 1, 2, . . . , N; and t ¼ 1, 2, . . . , T, where F denotes an m £ m matrix of slope
coef?cients, m
i
is an m £ 1 vector of individual-speci?c effects, 1
it
is an m £ 1 vector of
disturbances, and I
m
denotes the identity matrix of dimension m £ m. For simplicity,
we restrict our exposition to ?rst-order PVAR models.
Available literature mentions quite a few methods for determining the weights of
the variables. Mostly, these are econometric estimations with a macroeconomic model,
a reduced form aggregate demand function (backward looking IS curve), or a VAR
model. This study opines that the weights can also be determined by way of economic
arguments, such as the signi?cance of the variable for the banking system. This study,
on the other hand, feels that every variable in the index can be given equal weight.
Some studies employ the combination of above methods (Goodhart and Hofmann,
2001; Gauthier et al., 2004). The weighting factors are calculated by summing the
coef?cients of the variables and expressing them as a ratio (Montagnoli and
Napolitano, 2004):
Weighted variable X
i
ðW
i
Þ ¼
S Coefficient X
i;t;...n
S jCoefficient X
i...n;t;...n
j
By this approach, both the importance of the parameters of banking system and the
changes of its composition are duly taken into account for the analysis.
Accordingly:
Index Variable ¼
Weighted variable X
t
2X
t21
Weighted variable X
t
In our model of banking stability, we hypothesise that liquidity coverage ratio (LCR) is
dependent on net non-performing assets (NNPA) to net advances (NA), capital
adequacy ratio (CAR) and return on assets (ROA) and include these as the “variables of
interest”. Our hypothesis and selection of variables of interest is based on the available
literature for a banking dominated ?nancial system and is comparable to that of the
RBI methodology recently published (after this paper was submitted for publication) in
the working paper (Mishra et al., 2013). However, our approach differs slightly by not
considering “ef?ciency” (i.e. cost to income (operating expenses to income minus
interest expenses)) as one of the variables of interest in modelling banking stability
measures. Our standpoint is that “ef?ciency” is a determinant that has strong
multicollinearity with the pro?tability measure – “return on assets” and hence its
inclusion would lead to distortion in the model. We estimate a VAR model for banking
stability measures with the above detailed variables of interest.
Accordingly:
WLCR ¼ f{WNPA; WCAR; WROA} ð3Þ
Rewriting equation (2):
WLCR
it
¼ C þ WNPA
it
þ WCAR
it
þ WROA
it
þ1
it
ð4Þ
In the ensuing section, we present the analysis and the results of the econometric
analysis.
Banking stability
measures
33
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
4. Analysis and results
This study considers the core set of soundness indicators for the construction of the
index keeping in view the applicability of these determinants to the deposit taking
institutions (banking sector) in Indian ?nancial system. Capital adequacy measured
by regulatory capital to risk-weighted assets measures the strength of the banking
system in terms of capital adequacy to sustain the challenges of adverse impacts of
crisis like situations. “Nonperforming assets to total advances” represent asset
quality, earnings and pro?tability measures are represented by “return on assets”
and liquidity is measured by “liquidity coverage ratio”. We ?rst present the VAR
estimates of the variables in Table II and the comovement of the covariates in
Figure 3.
WLCR WCAR WNPA WROA
WLCR(21) 0.779497 20.180888 20.264254 0.079174
(0.04115) (0.17072) (0.09425) (0.02015)
[18.9428] [21.05955] [22.80366] [3.92924]
WLCR(22) 20.255397 20.514800 20.364963 20.105469
(0.04010) (0.16636) (0.09185) (0.01964)
[26.36910] [23.09443] [23.97361] [25.37133]
WCAR(21) 0.090358 0.701478 0.138745 0.006088
(0.01015) (0.04210) (0.02324) (0.00497)
[8.90536] [16.6639] [5.97000] [1.22531]
WCAR(22) 20.012079 0.249520 0.026662 0.012025
(0.01017) (0.04220) (0.02330) (0.00498)
[21.18750] [5.91272] [1.14440] [2.41428]
WNPA(21) 20.072334 0.220239 0.711947 0.016204
(0.01697) (0.07042) (0.03888) (0.00831)
[24.26151] [3.12748] [18.3123] [1.94952]
WNPA(22) 0.120170 20.006570 0.155808 20.018665
(0.01588) (0.06588) (0.03637) (0.00778)
[7.56727] [20.09972] [4.28360] [22.40034]
WROA(21) 0.529136 3.124731 20.459356 0.761692
(0.08226) (0.34127) (0.18841) (0.04028)
[6.43259] [9.15611] [22.43804] [18.9100]
WROA(22) 20.655473 21.597008 20.207827 0.087019
(0.08776) (0.36409) (0.20101) (0.04297)
[27.46902] [24.38628] [21.03392] [2.02496]
C 0.001560 0.006596 0.002476 0.001105
(0.00068) (0.00283) (0.00156) (0.00033)
[2.28889] [2.33266] [1.58579] [3.31174]
R
2
0.971583 0.977988 0.961263 0.932060
Adj. R
2
0.971240 0.977723 0.960796 0.931240
F-statistic 2,833.512 3,682.196 2,056.552 1,136.952
Log likelihood 1,894.357 938.2267 1,337.430 2,374.175
Akaike AIC 25.611180 22.765556 23.953662 27.039213
Schwarz SC 25.550775 22.705150 23.893257 26.978807
Notes: Sample (adjusted): 1998-2009; included observations: 672 after adjustments; standard errors in
parentheses and t-statistics in square brackets
Table II.
VAR estimates
JFEP
6,1
34
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
As the standard practice in VAR, analysis is to report results from Granger-causality
tests, impulse responses, and forecast error variance decompositions we present here
below the results of the econometric analyses. First, lag length selection criteria
determines the VAR model. We select the best lag length for the VAR model employing
the LR test on which Granger causality is based (Table III).
Correlation does not necessarily imply causation in any meaningful sense of that
word. Granger causality measures precedence and information content but does not by
itself indicate causality in the more common use of the term. The null hypothesis is that
x does not Granger-cause y in the ?rst regression and that y does not Granger-cause
x in the second regression (Granger, 1969). Based on the results of the lag order
selection criterion test, we use a lag length of 6 in estimating the F-statistic and the
probability values. Granger-causality statistics examine whether lagged values of one
variable helps to predict another variable (Table IV).
According to the results of Table IV, the p-values for all the arguments of
Granger-causality are signi?cant and hence we reject the null hypothesis and we
conclude that the variables of the model are Granger cause bi-directionally and therefore
exists causality among the covariates.
Figure 3.
Comovement of the
covariates
0
–1
1
2
3
4
1
–
9
6
3
–
9
8
5
–
0
0
7
–
0
2
9
–
0
4
1
1
–
0
6
1
3
–
0
8
1
6
–
9
6
1
8
–
9
8
2
0
–
0
0
2
2
–
0
2
2
4
–
0
4
2
6
–
0
6
2
8
–
0
8
3
1
–
9
6
3
3
–
9
8
3
5
–
0
0
3
7
–
0
2
3
9
–
0
4
4
1
–
0
6
4
3
–
0
8
4
6
–
9
6
4
8
–
9
8
5
0
–
0
0
5
2
–
0
2
5
4
–
0
4
5
6
–
0
6
WLCR WCAR
WNPA WROA
Lag LogL LR FPE AIC SC HQ
0 2,974.600 NA 2.04 £ 10
211
213.26161 213.22496 213.24716
1 4,671.900 3,356.713 1.12 £ 10
214
220.76741 220.58416 220.69517
2 4,772.759 197.6663 7.70 £ 10
215
221.14625 220.81640
a
221.01622
3 4,782.545 19.00449 7.92 £ 10
215
221.11851 220.64206 220.93069
4 4,833.983 98.97188 6.76 £ 10
215
221.27671 220.65366 221.03110
5 4,873.710 75.72918 6.08 £ 10
215
221.38263 220.61299 221.07923
6 4,948.348 140.9455
a
4.68 £ 10
215a
221.64441
a
220.72816 221.28322
a
Notes:
a
Indicates lag order selected by the criterion; LR – sequential modi?ed LR test statistic (each
test at 5 percent level); FPE – ?nal prediction error; AIC – Akaike information criterion; SC – Schwarz
information criterion; HQ – Hannan-Quinn information criterion
Table III.
VAR lag order
selection criteria
Banking stability
measures
35
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Robustness tests
We perform multivariate LM test to test the presence of autocorrelations and the
VAR residual portmanteau tests and for autocorrelations to establish the residual
autocorrelations. Further, we also perform the VAR Granger causality/block exogeneity
Wald tests, residual normality tests, and VAR residual heteroskedasticity tests with
without cross terms. However, in the interest of space the results are not presented here.
Impulse responses
A shock to the ith variable not only directly affects the ith variable but is also
transmitted to all of the other endogenous variables through the dynamic (lag)
structure of the VAR. An impulse response function traces the effect of a one-time
shock to one of the innovations on current and future values of the endogenous
variables (Figure 4). The accumulated response is the accumulated sum of the impulse
responses (Figure 5). It can be interpreted as the response to step impulse where the
same shock occurs in every period from the ?rst.
The impulse responses for the recursive VAR, ordered WLCR, WCAR, WNPA, and
WROA are plotted in Figure 4. The ?rst row shows the effect of an unexpected
1 percentage point increase in WLCR on all other three variables, as it works through
the recursive VAR system with the coef?cients estimated from actual data. The second
row shows the effect of an unexpected increase of 1 percentage point in the WCAR on
other three variables. Similarly, the third and fourth rows show the corresponding
effect for WNPA and WROA. Also plotted are ^1 standard error bands for each of the
impulse responses. These estimated impulse responses show patterns of persistent
common variation. For example (in the ?rst row of Figure 5), an unexpected rise in
WLCR slowly stabilizes at a level of around 4 percent, and is associated with a
persistent increase in WCAR (about 6 percent) and a moderate increases in WNPA
(about 2 percent) and WROA (about 1.5 percent). Residuals of the covariates are
presented in Figure 6.
The impulse responses observed in the analysis establish that LCR, CAR, NPA and
ROA are interrelated and can explain the banking system stability in the context of an
emerging economy banking stability.
Null hypothesis F-statistic Prob.
WNPA does not Granger cause WLCR 50.0629 6.0 £ 10
221
WLCR does not Granger cause WNPA 6.99518 0.0010
WCAR does not Granger cause WLCR 80.0327 7.0 £ 10
232
WLCR does not Granger cause WCAR 12.6460 4.0 £ 10
26
WROA does not Granger cause WLCR 40.9565 2.0 £ 10
217
WLCR does not Granger cause WROA 37.2950 4.0 £ 10
216
WCAR does not Granger cause WNPA 7.61499 0.0005
WNPA does not Granger cause WCAR 10.0468 5.0 £ 10
25
WROA does not Granger cause WNPA 0.74036 0.4773
WNPA does not Granger cause WROA 32.3571 4.0 £ 10
214
WROA does not Granger cause WCAR 33.2965 2.0 £ 10
214
WCAR does not Granger cause WROA 29.0865 8.0 £ 10
213
Notes: Sample: 1996-2009; lags: 2
Table IV.
Granger-causality
statistics
JFEP
6,1
36
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 4.
Impulse responses of
WLCR, WNPA, WCAR
and WROA using
recursive VAR
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
1
2
3
4
5
6
7
8
9
1
0
W
L
C
R
t
o
W
L
C
R
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
1
2
3
4
5
6
7
8
9
1
0
W
L
C
R
t
o
W
C
A
R
1
2
3
4
5
6
7
8
9
1
0
W
L
C
R
t
o
W
N
P
A
1
2
3
4
5
6
7
8
9
1
0
W
L
C
R
t
o
W
R
O
A
–
0
.
2 0
0
.
2
0
.
4
0
.
6
1
2
3
4
5
6
7
8
9
1
0
W
C
A
R
t
o
W
L
C
R
–
0
.
2 0
0
.
2
0
.
4
0
.
6
1
2
3
4
5
6
7
8
9
1
0
W
C
A
R
t
o
W
C
A
R
1
2
3
4
5
6
7
8
9
1
0
W
C
A
R
t
o
W
N
P
A
1
2
3
4
5
6
7
8
9
1
0
W
C
A
R
t
o
W
R
O
A
–
0
.
1 0
0
.
1
0
.
2
0
.
3
1
2
3
4
5
6
7
8
9
1
0
W
N
P
A
t
o
W
L
C
R
–
0
.
1 0
0
.
1
0
.
2
0
.
3
1
2
3
4
5
6
7
8
9
1
0
W
N
P
A
t
o
W
C
A
R
1
2
3
4
5
6
7
8
9
1
0
W
N
P
A
t
o
W
N
P
A
1
2
3
4
5
6
7
8
9
1
0
W
N
P
A
t
o
W
R
O
A
–
0
.
0
2
0
.
0
0
0
.
0
2
0
.
0
4
0
.
0
6
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
–
0
.
2 0
0
.
2
0
.
4
0
.
6
–
0
.
1 0
0
.
1
0
.
2
0
.
3
–
0
.
0
2
0
.
0
0
0
.
0
2
0
.
0
4
0
.
0
6
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
–
0
.
2 0
0
.
2
0
.
4
0
.
6
–
0
.
1 0
0
.
1
0
.
2
0
.
3
–
0
.
0
2
0
.
0
0
0
.
0
2
0
.
0
4
0
.
0
6
1
2
3
4
5
6
7
8
9
1
0
W
R
O
A
t
o
W
L
C
R
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
1
2
3
4
5
6
7
8
9
1
0
W
R
O
A
t
o
W
C
A
R
1
2
3
4
5
6
7
8
9
1
0
W
R
O
A
t
o
W
N
P
A
1
2
3
4
5
6
7
8
9
1
0
W
R
O
A
t
o
W
R
O
A
Banking stability
measures
37
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Variance decomposition of recursive VAR
This is an alternative method to the impulse response functions for examining the
effects of shocks to the dependent variables. This technique determines how much of
the forecast error variance for any variable in a system, is explained by innovations to
each explanatory variable, over a series of time horizons. Usually own series shocks
explain most of the error variance, although the shock will also affect other variables
in the system. It is also important to consider the ordering of the variables when
conducting these tests, as in practice the error terms of the equations in the VAR will
be correlated, so the result will be dependent on the order in which the equations are
estimated in the model.
While impulse response functions trace the effects of a shock to one endogenous
variable on to the other variables in the VAR, variance decomposition separates
the variation in an endogenous variable into the component shocks to the VAR. The
variance decomposition provides information about the relative importance of each
random innovation in affecting the variables in the VAR. Table V displays separate
variance decomposition for each endogenous variable. The second column, labeled
“SE”, contains the forecast error of the variable at the given forecast horizon.
The source of this forecast error is the variation in the current and future values of the
innovations to each endogenous variable in the VAR. The remaining columns give
the percentage of the forecast variance due to each innovation, with each row adding
up to 100. With the impulse responses, the variance decomposition based on the
Cholesky factor can change dramatically if the ordering of the variables in the VAR are
altered. For example, the ?rst period decomposition for the ?rst variable in the VAR
ordering is completely due to its own innovation.
The above presents results suggest considerable interaction among the variables.
For example, at period 10, 58.5 percent of the error in the forecast of liquidity is
attributed to capital adequacy (42 percent), asset quality (12 percent), and pro?tability
(4.5 percent) shocks in the recursive VAR. Similarly, at period 10, 39.5 percent of the
error in the forecast of capital adequacy is attributed to liquidity (6 percent), asset
quality (4 percent), and pro?tability (18 percent) shocks in the recursive VAR. For asset
quality at same period, 22 percent of the error in the forecast is attributed to liquidity
(8 percent), capital adequacy (11 percent), and pro?tability (18 percent) shocks in the
recursive VAR. Finally, for Pro?tability at the same period, 36 percent of the error in
the forecast is attributed to liquidity (2 percent), capital adequacy (32 percent),
Figure 5.
Accumulated responses
of covariates and
their residuals
1 2 3 4 5 6 7 8 9 10
Accumulated Response of WLCR
to cholesky one S.D. innovations
–0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
WLCR WNPA
WCAR WROA
1 2 3 4 5 6 7 8 9 10
Accumulated response of RESID01
to cholesky one S.D. innovations
–0.008
–0.004
0
0.004
0.008
0.012
0.016
RESID01 RESID02
RESID03 RESID04
JFEP
6,1
38
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 6.
Residuals of the covariates
–
0
.
1
6
–
0
.
1
2
–
0
.
0
8
–
0
.
0
4 0
0
.
0
4
0
.
0
8
0
.
1
2
0
.
1
6
1 – 9 8
3 – 9 9
5 – 0 0
7 – 0 1
9 – 0 2
1 1 – 0 3
1 3 – 0 4
1 5 – 0 5
1 7 – 0 6
1 9 – 0 7
2 1 – 0 8
2 3 – 0 9
2 6 – 9 8
2 8 – 9 9
3 0 – 0 0
3 2 – 0 1
3 4 – 0 2
3 6 – 0 3
3 8 – 0 4
4 0 – 0 5
4 2 – 0 6
4 4 – 0 7
4 6 – 0 8
4 8 – 0 9
5 1 – 9 8
5 3 – 9 9
5 5 – 0 0
W
L
C
R
R
e
s
i
d
u
a
l
s
–
0
.
4
–
0
.
2 0
0
.
2
0
.
4
0
.
6
0
.
8
1 – 9 8
3 – 9 9
5 – 0 0
7 – 0 1
9 – 0 2
1 1 – 0 3
1 3 – 0 4
1 5 – 0 5
1 7 – 0 6
1 9 – 0 7
2 1 – 0 8
2 3 – 0 9
2 6 – 9 8
2 8 – 9 9
3 0 – 0 0
3 2 – 0 1
3 4 – 0 2
3 6 – 0 3
3 8 – 0 4
4 0 – 0 5
4 2 – 0 6
4 4 – 0 7
4 6 – 0 8
4 8 – 0 9
5 1 – 9 8
5 3 – 9 9
5 5 – 0 0
W
C
A
R
R
e
s
i
d
u
a
l
s
–
0
.
3
–
0
.
2
–
0
.
1 0
0
.
1
0
.
2
0
.
3
0
.
4
1 – 9 8
3 – 9 9
5 – 0 0
7 – 0 1
9 – 0 2
1 1 – 0 3
1 3 – 0 4
1 5 – 0 5
1 7 – 0 6
1 9 – 0 7
2 1 – 0 8
2 3 – 0 9
2 6 – 9 8
2 8 – 9 9
3 0 – 0 0
3 2 – 0 1
3 4 – 0 2
3 6 – 0 3
3 8 – 0 4
4 0 – 0 5
4 2 – 0 6
4 4 – 0 7
4 6 – 0 8
4 8 – 0 9
5 1 – 9 8
5 3 – 9 9
5 5 – 0 0
W
N
P
A
R
e
s
i
d
u
a
l
s
–
0
.
0
6
–
0
.
0
4
–
0
.
0
2 0
0
.
0
2
0
.
0
4
1 – 9 8
3 – 9 9
5 – 0 0
7 – 0 1
9 – 0 2
1 1 – 0 3
1 3 – 0 4
1 5 – 0 5
1 7 – 0 6
1 9 – 0 7
2 1 – 0 8
2 3 – 0 9
2 6 – 9 8
2 8 – 9 9
3 0 – 0 0
3 2 – 0 1
3 4 – 0 2
3 6 – 0 3
3 8 – 0 4
4 0 – 0 5
4 2 – 0 6
4 4 – 0 7
4 6 – 0 8
4 8 – 0 9
5 1 – 9 8
5 3 – 9 9
5 5 – 0 0
W
R
O
A
R
e
s
i
d
u
a
l
s
Banking stability
measures
39
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
and asset quality (1 percent) shocks in the recursive VAR. These observations
illustrate and underscore the long-run impact of the volatilities in the variables of
banking stability.
Forecasting
State-of-the-art VAR forecasting systems contain more than three variables and allow
for time-varying parameters to capture important drifts in coef?cients (Sims, 1993).
Multistep ahead forecasts, computed by iterating forward the recursive VAR, are
assessed in Table VI.
The ?rst two forecast error statistics largely depend on the scale of the dependent
variable and are used as relative measures to compare forecasts for the same series
across different models; the smaller the error, the better the forecasting ability of
that model according to that criterion. Very low scores of root mean squared error
(RMSE) and mean absolute error (MAE) for the forecasts indicate the strength and
accuracy of the forecast based on the VAR model. The RMSE is computed using the
formula:
RMSE ¼
?????????????????????
P
ð y 2 ^ yÞ
2
n 2k 21
s
¼
????????????????????
RSS
n 2k 21
r
Period SE WLCR WCAR WNPA WROA
Variance decomposition of WLCR
1 0.014535 100.0000 0.000000 0.000000 0.000000
4 0.026667 72.61941 21.00196 2.486288 3.892338
6 0.030266 57.23783 32.01421 7.160574 3.587378
8 0.033300 47.41230 38.09176 10.85899 3.636939
10 0.035722 41.47237 41.72264 12.37306 4.431924
Variance decomposition of WCAR
1 0.060303 21.03585 78.96415 0.000000 0.000000
4 0.110671 13.49430 74.52642 1.480222 10.49906
6 0.134815 9.291467 74.32057 3.195048 13.19292
8 0.155314 7.019326 73.10044 4.095404 15.78483
10 0.173355 5.685542 71.53319 4.259851 18.52141
Variance decomposition of WNPA
1 0.033293 10.45330 4.10 £ 10
25
89.54666 0.000000
4 0.055051 6.097703 7.514617 84.50565 1.882026
6 0.064083 6.688434 9.529185 80.91078 2.871603
8 0.068821 7.650106 10.48979 78.81341 3.046687
10 0.070980 8.136431 11.19790 77.72246 2.943210
Variance decomposition of WROA
1 0.007118 0.592220 6.053441 3.433475 89.92086
4 0.012181 4.456874 16.09711 1.998224 77.44779
6 0.014187 3.376669 23.49209 1.480018 71.65122
8 0.015846 2.709682 28.77073 1.206576 67.31302
(1.40375) (5.02008) (0.67516) (5.30725)
10 0.017332 2.269778 32.50345 1.022762 64.20401
Notes: Cholesky ordering: WLCR, WCAR, WNPA, WROA; SE – Monte Carlo (100 repetitions)
Table V.
Results of variance
decomposition analysis
JFEP
6,1
40
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The remaining two statistics are scale invariant. The Theil inequality coef?cient
always lies between 0 and 1, where 0 indicates a perfect ?t. Further, as the ultimate
test of a forecasting model is its out-of-sample performance, Table VI focuses on
pseudo out-of-sample forecasts[1] over the period 1996-2009 (Figure 7).
5. Conclusion
We provide in this study an empirical evidence for the centrality of banking system
stability for aiding ?nancial stability in the context of banking dominated emerging
economy. Employing the bank speci?c variables of banking stability namely,
liquidity, asset quality, capital adequacy and pro?tability, this study has made two
valuable contributions. First, it has analysed the banking system stability by
employing VAR technique and has established the interdependence and comovement
of the banking stability covariates to the satisfaction of economic logic. Second, this
study is unique among the evolving body of literature that underscores the
signi?cant relationship between banking system resilience and ?nancial stability.
Further, the study has enabled us to understand that the ?nancial system and more
speci?cally the banking system in India has demonstrated continued stability
compared to other countries.
A well-functioning banking system is essential to sustain economic growth, both
to prop up the economic activities in the short run and to allocate resources ef?ciently
over the longer run. Capital market capitalization also helps strengthening ?nancial
stability. The overall approach to sustain ?nancial stability has to be multi-pronged.
Ensuring overall macroeconomic balance, enhancement in the macro-prudential
functioning of institutions and markets, and reinforcement of micro-prudential
institutional soundness through regulation and supervision need to be regularly
undertaken towards ?nancial stability. Financial markets are rapidly growing by
way of technology adoption, product innovation, and geographic and sectoral
integration. This swift development of ?nancial markets while contributing to
enhanced ?nancial stability may also throw up both bene?ts and new sources of
risks to ?nancial system.
Forecast statistics WLCRF WCARF WNPAF WROAF
Root mean squared error
a
0.020435 0.095363 0.094401 0.010598
Mean absolute error
b
0.009820 0.046692 0.050487 0.005676
Mean absolute percentage error 93.65728 38.29840 300.2287 73.03435
Theil inequality coef?cient 0.108798 0.105338 0.240491 0.179451
Bias proportion 0.000000 0.027040 0.013477 0.072535
Variance proportion 0.019645 0.044505 0.044977 0.003227
Covariance proportion 0.980355 0.928455 0.941546 0.924237
Notes:
a
The mean squared forecast error is computed as the average squared value of the forecast
error over the 1996-2009 out-of-sample period, and the resulting square root is the root mean squared
forecast error reported in the table; root mean squared errors (RMSEs) are the errors squared before
they are averaged and give a relatively high weight to large errors, which infers that RMSE is most
useful when large errors are particularly undesirable;
b
mean absolute error (MAE), which is a linear
score (that all the individual differences are weighted equally in the average), measures the magnitude
of the errors in a set of forecasts without considering their direction and measures accuracy for
continuous variable; entries are the root mean square error of forecasts computed recursively for VARs
Table VI.
Forecast statistics for
the variables
Banking stability
measures
41
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 7.
Forecasting
for the covariates
–
0
.
1 0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
1 – 9 6
3 – 9 8
5 – 0 0
7 – 0 2
9 – 0 4
1 1 – 0 6
1 3 – 0 8
1 6 – 9 6
1 8 – 9 8
2 0 – 0 0
2 2 – 0 2
2 4 – 0 4
2 6 – 0 6
2 8 – 0 8
3 1 – 9 6
3 3 – 9 8
3 5 – 0 0
3 7 – 0 2
3 9 – 0 4
4 1 – 0 6
4 3 – 0 8
4 6 – 9 6
4 8 – 9 8
5 0 – 0 0
5 2 – 0 2
5 4 – 0 4
5 6 – 0 6
W
L
C
R
F
±
2
S
.
E
.
W
C
A
R
F
±
2
S
.
E
.
W
N
P
A
F
±
2
S
.
E
.
W
R
O
A
F
±
2
S
.
E
.
–
1 0 1 2 3 4
1 – 9 6
3 – 9 8
5 – 0 0
7 – 0 2
9 – 0 4
1 1 – 0 6
1 3 – 0 8
1 6 – 9 6
1 8 – 9 8
2 0 – 0 0
2 2 – 0 2
2 4 – 0 4
2 6 – 0 6
2 8 – 0 8
3 1 – 9 6
3 3 – 9 8
3 5 – 0 0
3 7 – 0 2
3 9 – 0 4
4 1 – 0 6
4 3 – 0 8
4 6 – 9 6
4 8 – 9 8
5 0 – 0 0
5 2 – 0 2
5 4 – 0 4
5 6 – 0 6
–
0
.
5 0
0
.
5
1
.
0
1
.
5
2
.
0
2
.
5
1 – 9 6
3 – 9 8
5 – 0 0
7 – 0 2
9 – 0 4
1 1 – 0 6
1 3 – 0 8
1 6 – 9 6
1 8 – 9 8
2 0 – 0 0
2 2 – 0 2
2 4 – 0 4
2 6 – 0 6
2 8 – 0 8
3 1 – 9 6
3 3 – 9 8
3 5 – 0 0
3 7 – 0 2
3 9 – 0 4
4 1 – 0 6
4 3 – 0 8
4 6 – 9 6
4 8 – 9 8
5 0 – 0 0
5 2 – 0 2
5 4 – 0 4
5 6 – 0 6
–
0
.
1
0
–
0
.
0
5
0
.
0
0
0
.
0
5
0
.
1
0
0
.
1
5
0
.
2
0
0
.
2
5
1 – 9 6
3 – 9 8
5 – 0 0
7 – 0 2
9 – 0 4
1 1 – 0 6
1 3 – 0 8
1 6 – 9 6
1 8 – 9 8
2 0 – 0 0
2 2 – 0 2
2 4 – 0 4
2 6 – 0 6
2 8 – 0 8
3 1 – 9 6
3 3 – 9 8
3 5 – 0 0
3 7 – 0 2
3 9 – 0 4
4 1 – 0 6
4 3 – 0 8
4 6 – 9 6
4 8 – 9 8
5 0 – 0 0
5 2 – 0 2
5 4 – 0 4
5 6 – 0 6
JFEP
6,1
42
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Note
1. Pseudo out-of-sample forecasts are often referred to as pseudo or “simulated” out-of-sample
forecasts to emphasise that they simulate how these forecasts would have been computed in
real time, although of course this exercise is conducted retrospectively, not in real time.
References
Aka, B.E. (2006), “On the duration of the ?nancial system stability under liberalization”,
Emerging Markets Review, Vol. 7 No. 2, pp. 147-161.
Allen, W.A. and Wood, G. (2006), “De?ning and achieving ?nancial stability”, Journal of
Financial Stability, Vol. 2 No. 2, pp. 152-172.
Anzoategui, D., Soledad, M., Per? ´a, M. and Melecky, M. (2012), “Bank competition in Russia: an
examination at different levels of aggregation”, Emerging Markets Review, Vol. 13, pp. 42-57.
Beck, T., Cull, R. and Jerome, A.T. (2005), “Bank privatization and performance: evidence from
Nigeria”, Journal of Banking & Finance, Vol. 29, pp. 2355-2379.
Beck, T., Hesse, H., Kick, T. and Von Westernhagen, N. (2009), “Bank ownership and stability:
evidence from Germany”, Bundesbank Working Paper Series.
Blavarg, M. and Nimander, P. (2002), “Interbank exposures and systemic risk”, Proceedings of
the Third Joint Central Bank Research Conference on Risk Measurement and Systemic Risk,
Bank for International Settlements, Basel, pp. 287-305.
Borio, C. (2003), “Towards a macro prudential framework for ?nancial supervision and
regulation?”, BIS Working Papers No. 128, Basel.
Borio, C. and Lowe, P. (2002), “Asset prices, ?nancial and monetary stability: exploring the
nexus”, BIS Working Papers No. 114, Basel.
Boyd, J.H. and de Nicolo, G. (2005), “The theory of bank risk-taking and competition revisited”,
Journal of Finance, Vol. 60 No. 3, pp. 1329-1343.
Boyd, J.H. and Runkle, D.E. (1993), “Size and performance of banking ?rms: testing the
predictions of theory”, Journal of Monetary Economics, Vol. 31, pp. 47-67.
Calomiris, C.W. and Mason, J.R. (1997), “Contagion and bank failures during the great
depression: the June 1932 Chicago banking panic”, American Economic Review, Vol. 87
No. 5, pp. 863-883.
Calomiris, C.W. and Mason, J.R. (2000), “Causes of bank distress during the great depression”,
National Bureau of Economic Research Working Paper No. 7919, Cambridge, MA.
Crockett, A. (1997), “Maintaining ?nancial stability in a global economy”, paper presented at the
Federal Reserve Bank of Kansas City’s Symposium, Jackson Hole, WY, 28-30 August.
Davis, E.P. (1993), “Bank credit risk”, Working Paper Series No. 8, Vol. 8, Bank of England, London.
Davis, E.P. (2003), “Towards a typology for systemic ?nancial instability”, Public Policy
Discussion Papers No. 03-20, Economics and Finance Section, School of Social Sciences,
Brunel University, Uxbridge.
De Bandt, O. and Hartmann, P. (2002), “Systemic risk in banking: a survey”, in Goodhart, C. and
Illing, G. (Eds), Financial Crises, Contagion, and the Lender of Last Resort – A Reader,
Oxford University Press, London, pp. 249-298.
de Nicolo, G. (2000), “Size, charter value, and risk in banking: an international perspective”,
International Finance Discussion Papers No. 689, Board of Governors of the Federal
Reserve System, Washington, DC.
Diamond, D.W. (1984), “Financial intermediation and delegated monitoring”, Review of Economic
Studies, Vol. 51, pp. 393-414.
Banking stability
measures
43
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Diamond, D.W. and Dybvig, P. (1983), “Bank runs, deposit insurance and liquidity”, Journal of
Political Economy, Vol. 91, pp. 401-419.
Diamond, D.W. and Rajan, R. (2001), “Liquidity risk, liquidity creation, and ?nancial fragility:
a theory of banking”, Journal of Political Economy, Vol. 109, pp. 287-327.
Foot, M. (2003), “Protecting ?nancial stability – how good are we at it?”, speech given at
University of Birmingham, 6 June.
Gadanecz, B. and Jayaram, K. (2009), “Measures of ?nancial stability – a review”, in Bank for
International Settlements (Ed.), Proceedings of the IFC Conference on “Measuring
Financial Innovation and Its Impact”, Basel, 26-27 August 2008, IFC Bulletins Chapters,
Vol. 31, Bank for International Settlements, Basel, pp. 365-380.
Gauthier, C., Graham, C. and Liu, Y. (2004), “Financial conditions indexes for Canada”, Bank of
Canada Working Paper No. 2004-22.
Goodhart, C.A.E. (1987), “Why do banks need a central bank”, Oxford Economic Papers, Vol. 39,
pp. 75-89.
Goodhart, C.A.E. and Hofmann, B. (2001), “Asset prices, ?nancial conditions, and the
transmission of monetary policy”, paper presented for the Conference on Asset Prices,
Exchange Rates, and Monetary Policy, Stanford University, 2-3 March.
Goodhart, C.A.E., Sunirand, P. and Tsomocos, D.P. (2004), “A model to analyse ?nancial fragility:
applications”, Journal of Financial Stability, Vol. 1, pp. 1-35.
Goodhart, C.A.E., Sunirand, P. and Tsomocos, D.P. (2005), “A risk assessment model for banks”,
Annals of Finance, Vol. 1, pp. 197-224.
Goodhart, C.A.E., Sunirand, P. and Tsomocos, D.P. (2006a), “A model to analyse ?nancial
fragility”, Economic Theory, Vol. 27, pp. 107-142.
Goodhart, C.A.E., Sunirand, P. and Tsomocos, D.P. (2006b), “A time series analysis of ?nancial
fragility in the UK banking system”, Annals of Finance, Vol. 2, pp. 1-21.
Goodhart, C.A.E., Hartmann, P., Llewellyn, D., Rojas-Suarez, L. and Weisbrod, S. (1998),
Financial Regulation: Why, How and Where Now?, Routledge, London.
Gray, D.F., Merton, R.C. and Bodie, Z. (2007), “New framework for measuring and managing
macro-?nancial risk and ?nancial stability”, NBER Working Paper No. 13607.
Haldane, A., Hall, S., Saporta, V. and Tanaka, M. (2004), “Financial stability and macroeconomic
models”, Bank of England Financial Stability Review, June, pp. 80-88.
Hasan, I. and Dwyer, G. (1994), “Bank runs in the free banking period”, Journal of Money, Credit,
and Banking, Vol. 26, pp. 271-288.
Hawkins, J. and Klau, M. (2000), “Measuring potential vulnerabilities in emerging market
economies”, BIS Working Papers No. 91, October, Vol. 91.
Hellwig, M. (1991), “Banking, ?nancial intermediation, and corporate ?nance”, in Giovannini, A.
and Mayer, C. (Eds), European Financial Integration, Cambridge University Press,
Cambridge, pp. 35-63.
Hoggarth, G.A., Milne, A. and Wood, G.E. (1998), “Alternative routes to banking stability:
a comparison of UK and German banking systems”, Financial Stability Review, Vol. 5,
pp. 55-68.
Houben, A., Kakes, J. and Schinasi, G. (2004), “Toward a framework for safeguarding ?nancial
stability”, IMF Working Papers No. WP/04/101, International Capital Markets
Department, International Monetary Fund, Washington, DC.
IMF (2006), Financial Soundness Indicators: Compilation Guide, International Monetary Fund,
Washington, DC, March.
JFEP
6,1
44
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
IMF (2008), Global Financial Stability Report, International Monetary Fund, Washington, DC, April.
Issing, O. (2003), “Monetary and ?nancial stability: is there a trade-off?”, paper presented at the
Conference on Monetary Stability, Financial Stability and the Business Cycle, Bank for
International Settlements, Basel, 28-29 March.
Merton, R. (1974), “On the pricing of corporate debt: the risk structure of interest rates”, Journal
of Finance, Vol. 29, pp. 449-470.
Mishkin, F.S. (1994), “Global ?nancial instability: framework, events, issues”, Journal of
Economic Perspectives, Vol. 13, pp. 3-25.
Mishra, R.N., Majumdar, S. and Bhandia, D. (2013), “Banking stability – a precursor to ?nancial
stability”, Reserve Bank of India Working Paper Series No. WPS (DEPR): 01/2013, January.
Montagnoli, A. and Napolitano, O. (2004), “Financial condition index and interest rate settings:
a comparative analysis”, Money Macro and Finance Research Group Conference Series No. 1.
Morgan, J.P. (1999), CreditMetrics, 4th ed., available at: www.creditmetrics.com
Nelson, W.R. and Perli, R. (2005), “Selected indicators of ?nancial stability”, paper presented at
the 4th Joint Central Bank Research Conference on “Risk Measurement and Systemic
Risk”, ECB, Frankfurt am Main, November.
Pain, D. (2003), “The provisioning experience of the major UK banks: a small panel
investigation”, Working Paper No. 177, Bank of England, London.
Saunders, A. and Wilson, B. (1996), “Contagious bank runs: evidence from the 1929-33 period”,
Journal of Financial Intermediation, Vol. 5 No. 4, pp. 409-423.
Schinasi, G.J. (2004), “De?ning ?nancial stability”, IMF Working Paper No. WP/04/187,
International Capital Markets Department, International Monetary Fund, Washington, DC.
Sims, C. (1980), “Macroeconomics and reality”, Econometrica, Vol. 48 No. 1, pp. 1-48.
Tsomocos, D.P. (2003a), “Equilibrium analysis, banking, and ?nancial instability”, Journal of
Mathematical Economics, Vol. 39, pp. 619-655.
Tsomocos, D.P. (2003b), “Equilibrium analysis, banking, contagion and ?nancial fragility”,
Bank of England Working Paper Series No. 175.
Further reading
Bernanke, B. (2005), “The global saving glut and the US current account de?cit”, Homer Jones
Lecture, St Louis, MO, 14 April, available at: www.federalreserve.gov
Grossman, R. (1993), “The macroeconomic consequences of bank failures under the national
banking system”, Explorations in Economic History, Vol. 30, pp. 294-320.
Group of Ten (2001), Report on Consolidation in the Financial Sector, BIS, IMF, OECD, Basel.
Laeven, L. and Valencia, F. (2008), “Systemic banking crises: a new database”, IMF Working
Paper No. WP/08/224, International Monetary Fund, Washington, DC.
About the author
Dr Vighneswara Swamy is currently with IBS-Hyderabad as an Associate Professor in the
?nance area. His research interests include ?nancial stability, ?nancial intermediation, risk
management, ?nancial markets, banking and ?nance and development economics, etc. He has
published several papers in journals of international recognition as well as national-level
publications. Vighneswara Swamy can be contacted at: [email protected]
Banking stability
measures
45
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_816830893.pdf
This study aims to investigate the inter-relatedness and the dynamics of banking stability
measures and offers answers for some of the related issues such as does financial stability require
the soundness of banking institutions, the stability of markets, the absence of turbulence and low
volatility? and to what extent the soundness of banking sector in the case of emerging economies can
help financial system stability.
Journal of Financial Economic Policy
Testing the interrelatedness of banking stability measures
Vighneswara Swamy
Article information:
To cite this document:
Vighneswara Swamy , (2014),"Testing the interrelatedness of banking stability measures", J ournal of
Financial Economic Policy, Vol. 6 Iss 1 pp. 25 - 45
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-01-2013-0002
Downloaded on: 24 January 2016, At: 21:48 (PT)
References: this document contains references to 56 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 340 times since 2014*
Users who downloaded this article also downloaded:
Faten Ben Bouheni, (2014),"Banking regulation and supervision: can it enhance stability in Europe?",
J ournal of Financial Economic Policy, Vol. 6 Iss 3 pp. 244-269http://dx.doi.org/10.1108/J FEP-11-2013-0059
Eric Osei-Assibey, Baimba Augustine Bockarie, (2013),"Bank risks, capital and loan supply: evidence from
Sierra Leone", J ournal of Financial Economic Policy, Vol. 5 Iss 3 pp. 256-271http://dx.doi.org/10.1108/
J FEP-09-2012-0041
J ill M. Hendrickson, Mark W. Nichols, Daniel R. Fairchild, (2014),"Bank branch location and stability
during distress", J ournal of Financial Economic Policy, Vol. 6 Iss 2 pp. 133-151http://dx.doi.org/10.1108/
J FEP-07-2013-0026
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Testing the interrelatedness
of banking stability measures
Vighneswara Swamy
Department of Finance, IBS-Hyderabad, Hyderabad, India
Abstract
Purpose – This study aims to investigate the inter-relatedness and the dynamics of banking stability
measures and offers answers for some of the related issues such as does ?nancial stability require
the soundness of banking institutions, the stability of markets, the absence of turbulence and low
volatility? and to what extent the soundness of banking sector in the case of emerging economies can
help ?nancial system stability.
Design/methodology/approach – This study investigates banking stability by structuring a
recursive micro panel vector auto regressive (VAR) model and corroborates the signi?cance of the
interrelatedness of the bank-speci?c variables such as liquidity, asset quality, capital adequacy and
pro?tability by employing a robust panel data drawn from 56 leading banks for a period of 12 years.
Findings – A signi?cant contribution of this study is in establishing that liquidity in the
banking-dominated ?nancial system is reciprocally related with asset quality, capital adequacy, and
pro?tability of the banking system and in effectively forecasting banking stability employing micro
panel recursive VAR model.
Research limitations/implications – The study could be further broadened by employing a
macro and structural VAR modelling to forecast banking stability.
Practical implications – This paper is one among the evolving body of literature that underscores
the signi?cant relationship between banking system resilience and ?nancial stability in the context of
emerging economies dominated with banking systems. Further, the forecast model is able to capture
the dynamics of banking stability with greater and appreciable accuracy.
Originality/value – The uniqueness of the study is in modelling banking stability measures in the
context of banking-dominated emerging economy ?nancial systems by employing micro panel recursive
VAR model by deriving data from 58 leading banks for the period of 12 years from 1996 to 2009 and
in offering insights in understanding ?nancial stability with comprehensive literature review.
Keywords Indicator, Crisis, Financial stability, Banks and ?nancial institutions, Banking stability
Paper type Research paper
1. Introduction
A threat to ?nancial stability anywhere in the world is potentially a threat to ?nancial
stability everywhere. As ?nancial stability and macroeconomic stability are intricately
related, ?nancial stability can be vulnerable even if there is price stability and
macroeconomic stability and hence cannot be taken for granted. However, empirical
?ndings suggest that greater market power in the banking market results in higher
instability. Although banks are better capitalized in less competitive markets, their
default risk remains higher. Particularly in the context of banking dominated emerging
market ?nancial systems where banks dominate more than 70-80 percent of the ?nancial
system, banking stability assumes greater prominence in ensuring ?nancial stability.
This study seeks to test the interrelatedness of banking stability measures in the
context of the dynamics of banking stability and offers understanding about the
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JEL classi?cation – E44, E58, G1, G21, G28
Journal of Financial Economic Policy
Vol. 6 No. 1, 2014
pp. 25-45
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-01-2013-0002
Banking stability
measures
25
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
signi?cance of liquidity of the banking sector in ensuring banking stability in the case
of bank dominated ?nancial systems particularly in the case of emerging economies.
This paper is one among the evolving body of literature that underscores the
signi?cance of banking stability towards attaining ?nancial stability. This study
analyses the concept and de?nition of ?nancial stability and in that backdrop analyses
the banking stability in the case of emerging market economy like that of India. This
paper begins by presenting in Section 2, the theoretical framework signifying the
recent approaches on de?ning and analysing ?nancial stability placing the banking
system at the epicentre of analysis. We emphasise more speci?cally how banking
stability can augment ?nancial stability in emerging market economies like that of
India that successfully came out of the recent global ?nancial crisis. We present the
methodology involving the data and its sources and research design explaining the
empirical framework and estimation of the micro vector auto regressive (VAR) model
of banking stability in Section 3. The results of the analyses with discussion on the
?ndings are enunciated in Section 4, and the conclusion and policy implications are
offered in Section 5.
2. Banking stability – theoretical framework
Notwithstanding its extensive use, ?nancial stability is dif?cult to de?ne let alone
measure. A sound understanding of ?nancial stability necessitates a conceptual
framework (Houben et al., 2004). In understanding ?nancial stability, the ?rst limitation
has been yet there is no widely accepted model or analytical framework for assessing the
?nancial stability as this it is still in its infant stage of development and practice,
as compared with, for example, the analysis of monetary and/or macroeconomic stability
(Schinasi, 2004). The concept of ?nancial stability is nebulous with no commonly
accepted de?nition. However, there have been some attempts to de?ne ?nancial stability.
Houben et al. (2004) considering ?nancial stability as a continuum changeable overtime
and consistent with multiple combinations of its constituent elements, de?ne it as the
ability to help the economic system allocate resources, manage risks and absorb shocks.
The best approach according to Allen and Wood (2006) is to de?ne the characteristics of
an episode of ?nancial instability ?rst and then de?ne ?nancial stability as a state of
affairs in which episodes of instability are unlikely to occur. Davis (2003) identi?es three
generic types of ?nancial instability. The ?rst is centered on “bank failures”, typically
following loan or trading losses, the second involves extreme “market price volatility”
after a shift in expectations and the third being the one that is linked to the second,
involves protracted collapses of market liquidity and issuance.
Borio (2003) and others take a macroprudential viewpoint and state ?nancial
stability in terms of limiting risks of signi?cant real output losses associated with
episodes of ?nancial system-wide distress. Suggesting an information-based de?nition,
Mishkin (1994) states that ?nancial instability occurs when shocks to the ?nancial
system interfere with information ?ows so that the ?nancial system can no longer do
its job of channeling funds to those with productive investment opportunities. On the
other hand, Crockett (1997) proposes that ?nancial stability refers to the stability of key
institutions and markets that go to make up the ?nancial system. Further, Issing (2003)
and Foot (2003) have suggested that ?nancial stability is associated with ?nancial
market bubbles, or more generally, with volatility in ?nancial market proxies as these
bubbles impair ?nancial markets ef?ciency; however, in and of themselves, they do not
JFEP
6,1
26
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
constitute a de?ning characteristic of ?nancial fragility, and more generally ?nancial
instability. Suggesting institutionally oriented de?nitions, Haldane et al. (2004), among
others, have proposed that ?nancial instability could be de?ned as any deviation
from the optimal saving-investment plan of the economy that is due to ?nancial
imperfections in the ?nancial sector. Similarly, Goodhart et al. (2004, 2005, 2006a, b)
and Tsomocos (2003a, b) offer de?nitions for ?nancial stability that hinges upon the
welfare effects on the economy and distributional consequences arising during periods
of ?nancial instability.
The distinctiveness of the banking stability from the perspective of ?nancial
stability is a well-recognised idea in the literature. Banking assumes special status as
they are regarded as more vulnerable to instability than other sectors. Hellwig (1991)
takes a different view on the problem and states that banks are special because they
represent the availability of funds and governments want to exercise tighter control
over them. It also argued that banks assume signi?cance in view of their tendency
towards maturity mismatches between assets and liabilities that can expose them to
possibility of runs (Diamond and Dybvig, 1983; Diamond, 1984; Diamond and Rajan,
2001). Moreover, referring to the recent times, banking is involved in heavy interbank
lending and payment markets. These interlinkages, in the absence of safety net
provisions pose the threat of propagation of risk from one bank to the other, creating
interbank contagion risk – one form of systemic risk (De Bandt and Hartmann, 2002).
While the vulnerability to runs due to problems in the interbank market represents one
source of concerns about bank instability originating from the liability side of the
balance sheet, a second source of instability relates to bank risk-taking on the asset
side. Because of their substantial ?nancing from many small, relatively uninformed
depositors and an often-existing public safety net in response to the previously
mentioned vulnerability, banks can be prone to taking on “excessive” risk. Thus, the
speciality of banking system stability is a widely recognised idea in the ?nancial
stability literature (Goodhart, 1987; Goodhart et al., 1998). Thus, to sum up, banks are
closely linked with our everyday lives and activities. Any large disruption of banking
operations will affect society as a whole. Banking stability is, therefore, crucial to
minimising the extensive economic and social impact that may arise from problems in
the industry.
The issue of ?nancial stability is quite organically linked with banking stability.
Banking stability is a yardstick to determine whether an economy is adequately strong
enough to withstand both the internal and external shocks. On the other hand, ?nancial
stability is a by-product of stability conditions prevailing in banking system, ?nancial
markets, and the real economy and amongst them, banking stability appears to be a
vital ingredient to ?nancial stability. Banking stability in itself depends on the
ef?cacies of the several parameters of individual banks, e.g. asset quality, liquidity,
capital adequacy, and pro?tability, etc. Since, banking stability gets affected positively
or negatively with the prevailing conditions in the ?nancial market and the real
economy; ultimately, it determines as to what extent ?nancial stability is ensured in the
economy by its ability to absorb the shocks. As such, banking stability can be treated
as a forerunner of ?nancial stability in an economy. Accordingly, this paper takes into
consideration a constructive viewpoint and de?nes banking stability as a state of
affairs in which the ?nancial system can; achieve ef?cient allocation of resources;
assess and manage ?nancial risks; absorb the emerging shocks; ensure smooth
Banking stability
measures
27
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
payments and remittances; enhance equilibrium by managing asset and price
volatility; and lead the economy towards bene?ts of economic welfare.
2.1 Financial stability and banking sector
Banking sector is regarded as an important sector for the stability of ?nancial systems
as the banks play a major role in money creation, investment for economic growth,
?nance for businesses and households and in payment systems. Hence, faltering
banking systems are found to be associated with hyperin?ations and depressions in
the economic history. Global ?nancial crisis has drawn the attention of the policy
makers’ of advanced and emerging economies towards banking stability and placed it
on the top of their agenda. While some analysts view banking stability to be related in
part to size and ownership structures, some others observe point to the failure of
private banks as evidence of the fragility of short-term and pro?t-oriented banking.
A stable macroeconomic environment is essential for banking sector stability, mainly
because uncertainty about macroeconomic policies and wavering fundamentals, such
as economic growth and in?ation, renders it challenging for banks to assess credit
risks accurately. Subdued economic growth, due to macroeconomic uncertainty or for
other reasons, may impair bank soundness as it reduces the debt servicing capacity of
?rms and households. There is also a signi?cant debate on the implications of
competition for banking sector stability (Anzoategui et al., 2012) and new evidence
suggests that lack of competition can lead to fragility, especially if certain banks
become too big to fail. In an econometric attempt to deal with the issue of ?nancial
system stability in a framework of duration models, using macroeconomic data, Aka
(2006) observe that banking crises are likely to occur every ten years or so, even if the
?nancial system has been stable until the year before the crises. Indeed, the risk of
banking crises builds up during a calm period. The results also reveal that factors such
as fundamentals weakness, structural characteristics, contagion phenomenon and past
experiences with banking crises play an important role.
Particularly in emerging economies, at present, the banking sector is by far the most
important part of the ?nancial system in all and is, therefore, also the main source of
risk for ?nancial stability. This is all the more so because the lack of well-functioning
equity markets confronts banks with relatively high credit risks, as bank credit is
necessary (to some extent) to substitute for equity. One of the important sources of
vulnerability that can affect ?nancial stability and lead to a ?nancial crisis can be the
weakness (such as a high level of short-term debt) in the ?nancial structure of
the economy, i.e. the composition and the size of the assets and liabilities on the balance
sheet. A ?nancial crisis follows when the demand for ?nancial assets of one or more
sectors plummets and consequently the banking system fails to meet the out?ows or
may be unable to attract new ?nancing or roll over existing short-term liabilities.
In this direction, ?nancial soundness matters much during the ?nancial crisis because
it gives some indication of how likely it is that ?nancial problems would be transmitted
into the real economy.
2.2 Banking stability measures
Measurement of banking stability has some challenges as it faces complication in
assessing, in contrast to other elements of the ?nancial system such as; securities
values, interbank relationships that can be at the origin of bank contagion phenomena
JFEP
6,1
28
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
or the values of and correlations between loan portfolios are mostly hard to measure
and monitor. As observed by Blavarg and Nimander (2002), even central banks and
supervisory authorities usually do not have continuous information about interbank
exposures. Financial stability literature provides a number of quantitative measures in
order to assess ?nancial stability. IMF (2006) recommends 12 core ?nancial sector
indicators (FSIs) for deposit takers and are categorised as capital-based FSIs,
asset-based FSIs, and income and expense FSIs. Similarly, such indicators are used as
monitoring variables used in the studies by Hawkins and Klau (2000), Nelson and Perli
(2005) and Gray et al. (2007) which focus on market pressures, external vulnerability
and banking system vulnerability. Some authors have focused on three indicators of
?nancial stability, namely, the z-score as measure of distance from insolvency, the
NPL-score as indicator of lending risk, and the probability of distress score (PD-score)
as measure of actual insolvency risk (Beck et al., 2009).
The choice of employing bank equity prices for measuring banking stability is
observed to be motivated by Merton’s (1974) option-theoretic framework toward
default. This approach has become the cornerstone of a large body of approaches
towards quantifying credit risk and modelling credit rating migrations including
Morgan’s (1999) CreditMetrics. However, ?nancial stability literature ?nds a number
of approaches that do not rely on market indicators as banking stability measures.
While Saunders and Wilson (1996) and Calomiris and Mason (1997) study deposit
withdrawals as a measure of banking stability in their study of failing and non-failing
banks during the great depression, Hasan and Dwyer (1994) measure autocorrelation of
bank failures after controlling for macroeconomic fundamentals during various
episodes of US banking history. Further, Calomiris and Mason (2000) evaluate the
survival time of banks during the great depression, with explanatory variables
including national and regional macroeconomic variables, dummies for well-known
panics and the level of deposits in the same county (contagion effect).
Bank size is considered a measure of bank stability, and its relationship has been
ambiguous in the academic literature (Boyd and Runkle, 1993; Calomiris and Mason,
2000) while de Nicolo (2000) observes a negative and signi?cant relationship.
Moreover, a higher amount of risky assets is expected to be associated with higher
bank failure and too rapid expansion of loan portfolio can lean to greater risk taking by
banks (Pain, 2003; Davis, 1993; Borio and Lowe, 2002). Literature also records market
concentration as a measure of bank stability as Beck et al. (2005) observe that market
concentration is associated with more ?nancial stability. In contrast, Boyd and
de Nicolo (2005) suggest that systems that are more competitive are conducive to bank
stability. On the other hand, Hoggarth et al. (1998) ?nd that a less competitive German
banking system seems to be more stable in their comparison of German and UK banks.
Most of the central banks across the globe have been publishing their ?nancial
stability reports (FSRs) extensively focusing on various market segments and banking
related variables. Banking ratios are widely analysed in most of such reports although
there are some differences. Some reports seem to concentrate on the banks’ performance
and risks in considerable detail, while the others take account insurance, hedge funds
and other forms of nonbank ?nancial intermediation. However, where banking is the
main form of intermediation, the available information largely depends on the level of
supervisory input in preparing the report. Some central banks, compute a banking
stability index using weighted average of sub-indicators of banking sector soundness
Banking stability
measures
29
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
including capital adequacy, pro?tability, balance sheet liquidity, asset quality, credit and
currency risk. On the other hand, some central banks calculate a ?nancial strength index
as a weighted average of partial indicators of the ?nancial soundness of banks. The
index combines six areas of ?nancial soundness indicators, namely capital adequacy,
pro?tability, liquidity, asset quality, interest rate risk and exchange rate risk.
Developing a banking stability index, IMF considers the banking system as a
portfolio, and based on the market-based probability of default of each individual bank
it estimates the joint default probabilities, i.e. the expected number of bank defaults in
the system, given that at least one bank defaults (IMF, 2008). In the Indian context,
Reserve Bank of India (RBI) employs a banking stability map and banking stability
index to assess inherent risk factors that impact the stability of the banking sector. The
indicator is based on ?ve indices, which represent the ?ve dimensions namely,
soundness, asset-quality, pro?tability, liquidity, and ef?ciency. A composite measure
of each dimension is calculated as a weighted average of a set of standardized ratios
that are relevant in assessing the dimension. Based on the individual composite indices
for each dimension, the banking stability indicator (BSI) is constructed as a simple
average of the above ?ve composite sub-indices. The higher value of the indicator
would suggest lower stability (Mishra et al., 2013). Gadanecz and Jayaram (2009),
summarizing the FSIs state that the ?nancial sector is characterised by monetary
aggregates, real interest rates, risk measures for the banking sector, banks’ capital and
liquidity ratios, the quality of their loan book, standalone credit ratings and the
concentration/systemic focus of their lending activities. All these proxies can be
re?ective of problems in the banking or ?nancial sector and, if a crisis occurs, they can
gauge the cost of such a crisis to the real economy.
2.3 Financial soundness in Indian banking system
Banking sector is by far the most central part of the ?nancial system in most of the
emerging economies and is, therefore, also the main source of risk for ?nancial
stability. Undoubtedly, ?nancial soundness of banks has a signi?cant sway on the
stability of the ?nancial system as a whole as the banking system constitutes more
than 75 percent of the ?nancial markets in India. The Indian banking system endured
the onslaught of the global ?nancial crisis and a factor that bolstered the normal
functioning of the banking system even in the face of one of the largest global ?nancial
crisis was its robust capital adequacy. Further, the core banking sector indicators for
India like; capital adequacy ratio (CAR) – Tier-1, gross non-performing assets
(GNPAs) to total loans, net non-performing assets (NNPAs) to total loans and return on
equity (ROE) have experienced downward pressure during the recent recession period
(Figure 1). On the contrary, liquid assets to total assets ratio has moved upwards
indicating the tendency of the banks to hold cash during the times of recession instead
of investing in loans or investment products.
Under Basel II, capital to risk-weighted assets ratio (CRAR) of Indian banks as at
end-March 2009 was at 14 percent, far above the stipulated level of 9 percent (Figure 2).
This suggests that Indian banks have successfully managed to meet the increased
capital requirement under the amended framework.
Furthermore, between March 2009 and 2010, there was a surge by about
0.5 percentage point in the CRAR re?ecting further strengthening of their capital
adequacy under the new framework.
JFEP
6,1
30
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3. Methodology and data
In view of the fact that assessment of ?nancial stability in general is made on a
broad-spectrum of risk factors, one cannot expect a single model to capture
satisfactorily all the risk factors originating and developing inside and outside the
?nancial system, respectively. Instead, a suite of models may be required. However, the
objective of the ensuing segment of this paper is to analyse the salient parameters of
banking sector performance and behaviour and to establish the signi?cance of banking
sector stability in the context of bank-dominated ?nancial system of Indian economy
Figure 2.
Capital to risk-weighted
assets ratio-bank
group-wise
(as at end-March)
Note:
a
Includes IDBI Bank Ltd
0
2
4
6
8
10
12
14
16
18
20
Basel-1 2009 Basel-I 2010 Basel-II 2009 Basel-II 2010
Public sector banks Nationalised banks
a
SBI group Private sector banks
Old private sector banks
Source: Report on Trend and Progress of Banking in India
2009-2010 of RBI
Figure 1.
Core banking sector
indicators for India
Source: International Financial Statistics (IFS) of IMF
0
5
10
15
20
A:2008 Q:2009:1
a
Q:2009:3
a
Q:2010:1
a
Q:2010:3
a
I
n
P
e
r
c
e
n
t
Note:
a
Quarterly figures : A - annual Q - quarter
Car Car-Tier-1
Net NPAs to Capital Gross NPAs to Total Loans
Return on Equity
Banking stability
measures
31
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
by employing a micro model of VAR. The proposed micro VAR model involves the
most relevant parameters of banking stability namely, liquidity, asset quality, capital
adequacy and pro?tability.
Data and variables
The data for the analysis is sourced from the robust database of RBI. The variables
might be considerably adopted to measure the degree of volatility and soundness in
the banking sector and thereby in?uence the ?nancial stability is listed in Table I.
For this purpose, a panel data (annual) involving weightages for the variables of bank
performance and behaviour for the period from 1996 to 2009 covering 56 commercial
banks in India is constructed.
The model
Vector auto regressions (VARs) are powerful tools for describing data and for
generating reliable multivariate benchmark forecasts. Sims (1980) advocated VAR
models as providing a theory-free method to estimate economic relationships, thus
being an alternative to the “incredible identi?cation restrictions” in structural models.
Used wisely and based on economic reasoning and institutional detail, VARs both
can ?t the data and, at their best, can provide sensible estimates of some causal
connections. Although VARs have limitations when it comes to structural inference
and policy analysis, so do the alternatives. A recursive VAR constructs the error terms
in the each regression equation to be uncorrelated with the error in the preceding
equations. This is done by judiciously including some contemporaneous values as
regressors.
Let Y
it
be an m £ 1 vector of random variables for the ith cross-sectional unit at time
t, and suppose that the Y
it
’s are generated by the following panel vector autoregressive
model of order one (PVAR):
Y
it
¼ FY
i;t21
þ1
it
ð1Þ
Y
it
¼ ðI
m
2FÞm
i
þFY
i;t21
þ1
it
ð2Þ
Variable Description
Capital adequacy
ratio (CAR)
De?ned as the amount of regulatory capital to be maintained by a bank to
account for various risks inbuilt in the banking system
Capital Adequacy Ratio ¼
Total Capital ðTier I Capital þTier II CapitalÞ
Market Risk ðRWAÞ þCredit Risk ðRWAÞ þOperation Risk ðRWAÞ
RWA – risk weighted assets
The higher the ratio the better is for the bank’s stability
Return on assets
(ROA)
Return on Assets ¼ Net Pro?t/Assets ¼ (Net Pro?t/Total Income)
*
(Total
Income/Assets)
The higher the ratio the better is for the bank’s stability
Net non-performing
assets (NNPA) to
net advances (NA)
Net NPA to Net Advances ¼ Net NPA=Net Advances
The lower the ratio the better is for the bank’s stability
Management of non-performing assets is a key to the stability and continued
viability of the banking sector
Liquidity coverage
ratio (LCR)
Liquidity Coverage Ratio ¼
ðCash þSLRþother short term investmentsÞ
Short term liabilities
The lower the ratio indicates less liquidity
Table I.
Description of key
variables
JFEP
6,1
32
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
for i ¼ 1, 2, . . . , N; and t ¼ 1, 2, . . . , T, where F denotes an m £ m matrix of slope
coef?cients, m
i
is an m £ 1 vector of individual-speci?c effects, 1
it
is an m £ 1 vector of
disturbances, and I
m
denotes the identity matrix of dimension m £ m. For simplicity,
we restrict our exposition to ?rst-order PVAR models.
Available literature mentions quite a few methods for determining the weights of
the variables. Mostly, these are econometric estimations with a macroeconomic model,
a reduced form aggregate demand function (backward looking IS curve), or a VAR
model. This study opines that the weights can also be determined by way of economic
arguments, such as the signi?cance of the variable for the banking system. This study,
on the other hand, feels that every variable in the index can be given equal weight.
Some studies employ the combination of above methods (Goodhart and Hofmann,
2001; Gauthier et al., 2004). The weighting factors are calculated by summing the
coef?cients of the variables and expressing them as a ratio (Montagnoli and
Napolitano, 2004):
Weighted variable X
i
ðW
i
Þ ¼
S Coefficient X
i;t;...n
S jCoefficient X
i...n;t;...n
j
By this approach, both the importance of the parameters of banking system and the
changes of its composition are duly taken into account for the analysis.
Accordingly:
Index Variable ¼
Weighted variable X
t
2X
t21
Weighted variable X
t
In our model of banking stability, we hypothesise that liquidity coverage ratio (LCR) is
dependent on net non-performing assets (NNPA) to net advances (NA), capital
adequacy ratio (CAR) and return on assets (ROA) and include these as the “variables of
interest”. Our hypothesis and selection of variables of interest is based on the available
literature for a banking dominated ?nancial system and is comparable to that of the
RBI methodology recently published (after this paper was submitted for publication) in
the working paper (Mishra et al., 2013). However, our approach differs slightly by not
considering “ef?ciency” (i.e. cost to income (operating expenses to income minus
interest expenses)) as one of the variables of interest in modelling banking stability
measures. Our standpoint is that “ef?ciency” is a determinant that has strong
multicollinearity with the pro?tability measure – “return on assets” and hence its
inclusion would lead to distortion in the model. We estimate a VAR model for banking
stability measures with the above detailed variables of interest.
Accordingly:
WLCR ¼ f{WNPA; WCAR; WROA} ð3Þ
Rewriting equation (2):
WLCR
it
¼ C þ WNPA
it
þ WCAR
it
þ WROA
it
þ1
it
ð4Þ
In the ensuing section, we present the analysis and the results of the econometric
analysis.
Banking stability
measures
33
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
4. Analysis and results
This study considers the core set of soundness indicators for the construction of the
index keeping in view the applicability of these determinants to the deposit taking
institutions (banking sector) in Indian ?nancial system. Capital adequacy measured
by regulatory capital to risk-weighted assets measures the strength of the banking
system in terms of capital adequacy to sustain the challenges of adverse impacts of
crisis like situations. “Nonperforming assets to total advances” represent asset
quality, earnings and pro?tability measures are represented by “return on assets”
and liquidity is measured by “liquidity coverage ratio”. We ?rst present the VAR
estimates of the variables in Table II and the comovement of the covariates in
Figure 3.
WLCR WCAR WNPA WROA
WLCR(21) 0.779497 20.180888 20.264254 0.079174
(0.04115) (0.17072) (0.09425) (0.02015)
[18.9428] [21.05955] [22.80366] [3.92924]
WLCR(22) 20.255397 20.514800 20.364963 20.105469
(0.04010) (0.16636) (0.09185) (0.01964)
[26.36910] [23.09443] [23.97361] [25.37133]
WCAR(21) 0.090358 0.701478 0.138745 0.006088
(0.01015) (0.04210) (0.02324) (0.00497)
[8.90536] [16.6639] [5.97000] [1.22531]
WCAR(22) 20.012079 0.249520 0.026662 0.012025
(0.01017) (0.04220) (0.02330) (0.00498)
[21.18750] [5.91272] [1.14440] [2.41428]
WNPA(21) 20.072334 0.220239 0.711947 0.016204
(0.01697) (0.07042) (0.03888) (0.00831)
[24.26151] [3.12748] [18.3123] [1.94952]
WNPA(22) 0.120170 20.006570 0.155808 20.018665
(0.01588) (0.06588) (0.03637) (0.00778)
[7.56727] [20.09972] [4.28360] [22.40034]
WROA(21) 0.529136 3.124731 20.459356 0.761692
(0.08226) (0.34127) (0.18841) (0.04028)
[6.43259] [9.15611] [22.43804] [18.9100]
WROA(22) 20.655473 21.597008 20.207827 0.087019
(0.08776) (0.36409) (0.20101) (0.04297)
[27.46902] [24.38628] [21.03392] [2.02496]
C 0.001560 0.006596 0.002476 0.001105
(0.00068) (0.00283) (0.00156) (0.00033)
[2.28889] [2.33266] [1.58579] [3.31174]
R
2
0.971583 0.977988 0.961263 0.932060
Adj. R
2
0.971240 0.977723 0.960796 0.931240
F-statistic 2,833.512 3,682.196 2,056.552 1,136.952
Log likelihood 1,894.357 938.2267 1,337.430 2,374.175
Akaike AIC 25.611180 22.765556 23.953662 27.039213
Schwarz SC 25.550775 22.705150 23.893257 26.978807
Notes: Sample (adjusted): 1998-2009; included observations: 672 after adjustments; standard errors in
parentheses and t-statistics in square brackets
Table II.
VAR estimates
JFEP
6,1
34
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
As the standard practice in VAR, analysis is to report results from Granger-causality
tests, impulse responses, and forecast error variance decompositions we present here
below the results of the econometric analyses. First, lag length selection criteria
determines the VAR model. We select the best lag length for the VAR model employing
the LR test on which Granger causality is based (Table III).
Correlation does not necessarily imply causation in any meaningful sense of that
word. Granger causality measures precedence and information content but does not by
itself indicate causality in the more common use of the term. The null hypothesis is that
x does not Granger-cause y in the ?rst regression and that y does not Granger-cause
x in the second regression (Granger, 1969). Based on the results of the lag order
selection criterion test, we use a lag length of 6 in estimating the F-statistic and the
probability values. Granger-causality statistics examine whether lagged values of one
variable helps to predict another variable (Table IV).
According to the results of Table IV, the p-values for all the arguments of
Granger-causality are signi?cant and hence we reject the null hypothesis and we
conclude that the variables of the model are Granger cause bi-directionally and therefore
exists causality among the covariates.
Figure 3.
Comovement of the
covariates
0
–1
1
2
3
4
1
–
9
6
3
–
9
8
5
–
0
0
7
–
0
2
9
–
0
4
1
1
–
0
6
1
3
–
0
8
1
6
–
9
6
1
8
–
9
8
2
0
–
0
0
2
2
–
0
2
2
4
–
0
4
2
6
–
0
6
2
8
–
0
8
3
1
–
9
6
3
3
–
9
8
3
5
–
0
0
3
7
–
0
2
3
9
–
0
4
4
1
–
0
6
4
3
–
0
8
4
6
–
9
6
4
8
–
9
8
5
0
–
0
0
5
2
–
0
2
5
4
–
0
4
5
6
–
0
6
WLCR WCAR
WNPA WROA
Lag LogL LR FPE AIC SC HQ
0 2,974.600 NA 2.04 £ 10
211
213.26161 213.22496 213.24716
1 4,671.900 3,356.713 1.12 £ 10
214
220.76741 220.58416 220.69517
2 4,772.759 197.6663 7.70 £ 10
215
221.14625 220.81640
a
221.01622
3 4,782.545 19.00449 7.92 £ 10
215
221.11851 220.64206 220.93069
4 4,833.983 98.97188 6.76 £ 10
215
221.27671 220.65366 221.03110
5 4,873.710 75.72918 6.08 £ 10
215
221.38263 220.61299 221.07923
6 4,948.348 140.9455
a
4.68 £ 10
215a
221.64441
a
220.72816 221.28322
a
Notes:
a
Indicates lag order selected by the criterion; LR – sequential modi?ed LR test statistic (each
test at 5 percent level); FPE – ?nal prediction error; AIC – Akaike information criterion; SC – Schwarz
information criterion; HQ – Hannan-Quinn information criterion
Table III.
VAR lag order
selection criteria
Banking stability
measures
35
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Robustness tests
We perform multivariate LM test to test the presence of autocorrelations and the
VAR residual portmanteau tests and for autocorrelations to establish the residual
autocorrelations. Further, we also perform the VAR Granger causality/block exogeneity
Wald tests, residual normality tests, and VAR residual heteroskedasticity tests with
without cross terms. However, in the interest of space the results are not presented here.
Impulse responses
A shock to the ith variable not only directly affects the ith variable but is also
transmitted to all of the other endogenous variables through the dynamic (lag)
structure of the VAR. An impulse response function traces the effect of a one-time
shock to one of the innovations on current and future values of the endogenous
variables (Figure 4). The accumulated response is the accumulated sum of the impulse
responses (Figure 5). It can be interpreted as the response to step impulse where the
same shock occurs in every period from the ?rst.
The impulse responses for the recursive VAR, ordered WLCR, WCAR, WNPA, and
WROA are plotted in Figure 4. The ?rst row shows the effect of an unexpected
1 percentage point increase in WLCR on all other three variables, as it works through
the recursive VAR system with the coef?cients estimated from actual data. The second
row shows the effect of an unexpected increase of 1 percentage point in the WCAR on
other three variables. Similarly, the third and fourth rows show the corresponding
effect for WNPA and WROA. Also plotted are ^1 standard error bands for each of the
impulse responses. These estimated impulse responses show patterns of persistent
common variation. For example (in the ?rst row of Figure 5), an unexpected rise in
WLCR slowly stabilizes at a level of around 4 percent, and is associated with a
persistent increase in WCAR (about 6 percent) and a moderate increases in WNPA
(about 2 percent) and WROA (about 1.5 percent). Residuals of the covariates are
presented in Figure 6.
The impulse responses observed in the analysis establish that LCR, CAR, NPA and
ROA are interrelated and can explain the banking system stability in the context of an
emerging economy banking stability.
Null hypothesis F-statistic Prob.
WNPA does not Granger cause WLCR 50.0629 6.0 £ 10
221
WLCR does not Granger cause WNPA 6.99518 0.0010
WCAR does not Granger cause WLCR 80.0327 7.0 £ 10
232
WLCR does not Granger cause WCAR 12.6460 4.0 £ 10
26
WROA does not Granger cause WLCR 40.9565 2.0 £ 10
217
WLCR does not Granger cause WROA 37.2950 4.0 £ 10
216
WCAR does not Granger cause WNPA 7.61499 0.0005
WNPA does not Granger cause WCAR 10.0468 5.0 £ 10
25
WROA does not Granger cause WNPA 0.74036 0.4773
WNPA does not Granger cause WROA 32.3571 4.0 £ 10
214
WROA does not Granger cause WCAR 33.2965 2.0 £ 10
214
WCAR does not Granger cause WROA 29.0865 8.0 £ 10
213
Notes: Sample: 1996-2009; lags: 2
Table IV.
Granger-causality
statistics
JFEP
6,1
36
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 4.
Impulse responses of
WLCR, WNPA, WCAR
and WROA using
recursive VAR
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
1
2
3
4
5
6
7
8
9
1
0
W
L
C
R
t
o
W
L
C
R
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
1
2
3
4
5
6
7
8
9
1
0
W
L
C
R
t
o
W
C
A
R
1
2
3
4
5
6
7
8
9
1
0
W
L
C
R
t
o
W
N
P
A
1
2
3
4
5
6
7
8
9
1
0
W
L
C
R
t
o
W
R
O
A
–
0
.
2 0
0
.
2
0
.
4
0
.
6
1
2
3
4
5
6
7
8
9
1
0
W
C
A
R
t
o
W
L
C
R
–
0
.
2 0
0
.
2
0
.
4
0
.
6
1
2
3
4
5
6
7
8
9
1
0
W
C
A
R
t
o
W
C
A
R
1
2
3
4
5
6
7
8
9
1
0
W
C
A
R
t
o
W
N
P
A
1
2
3
4
5
6
7
8
9
1
0
W
C
A
R
t
o
W
R
O
A
–
0
.
1 0
0
.
1
0
.
2
0
.
3
1
2
3
4
5
6
7
8
9
1
0
W
N
P
A
t
o
W
L
C
R
–
0
.
1 0
0
.
1
0
.
2
0
.
3
1
2
3
4
5
6
7
8
9
1
0
W
N
P
A
t
o
W
C
A
R
1
2
3
4
5
6
7
8
9
1
0
W
N
P
A
t
o
W
N
P
A
1
2
3
4
5
6
7
8
9
1
0
W
N
P
A
t
o
W
R
O
A
–
0
.
0
2
0
.
0
0
0
.
0
2
0
.
0
4
0
.
0
6
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
–
0
.
2 0
0
.
2
0
.
4
0
.
6
–
0
.
1 0
0
.
1
0
.
2
0
.
3
–
0
.
0
2
0
.
0
0
0
.
0
2
0
.
0
4
0
.
0
6
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
–
0
.
2 0
0
.
2
0
.
4
0
.
6
–
0
.
1 0
0
.
1
0
.
2
0
.
3
–
0
.
0
2
0
.
0
0
0
.
0
2
0
.
0
4
0
.
0
6
1
2
3
4
5
6
7
8
9
1
0
W
R
O
A
t
o
W
L
C
R
–
0
.
0
2 0
0
.
0
2
0
.
0
4
0
.
0
6
1
2
3
4
5
6
7
8
9
1
0
W
R
O
A
t
o
W
C
A
R
1
2
3
4
5
6
7
8
9
1
0
W
R
O
A
t
o
W
N
P
A
1
2
3
4
5
6
7
8
9
1
0
W
R
O
A
t
o
W
R
O
A
Banking stability
measures
37
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Variance decomposition of recursive VAR
This is an alternative method to the impulse response functions for examining the
effects of shocks to the dependent variables. This technique determines how much of
the forecast error variance for any variable in a system, is explained by innovations to
each explanatory variable, over a series of time horizons. Usually own series shocks
explain most of the error variance, although the shock will also affect other variables
in the system. It is also important to consider the ordering of the variables when
conducting these tests, as in practice the error terms of the equations in the VAR will
be correlated, so the result will be dependent on the order in which the equations are
estimated in the model.
While impulse response functions trace the effects of a shock to one endogenous
variable on to the other variables in the VAR, variance decomposition separates
the variation in an endogenous variable into the component shocks to the VAR. The
variance decomposition provides information about the relative importance of each
random innovation in affecting the variables in the VAR. Table V displays separate
variance decomposition for each endogenous variable. The second column, labeled
“SE”, contains the forecast error of the variable at the given forecast horizon.
The source of this forecast error is the variation in the current and future values of the
innovations to each endogenous variable in the VAR. The remaining columns give
the percentage of the forecast variance due to each innovation, with each row adding
up to 100. With the impulse responses, the variance decomposition based on the
Cholesky factor can change dramatically if the ordering of the variables in the VAR are
altered. For example, the ?rst period decomposition for the ?rst variable in the VAR
ordering is completely due to its own innovation.
The above presents results suggest considerable interaction among the variables.
For example, at period 10, 58.5 percent of the error in the forecast of liquidity is
attributed to capital adequacy (42 percent), asset quality (12 percent), and pro?tability
(4.5 percent) shocks in the recursive VAR. Similarly, at period 10, 39.5 percent of the
error in the forecast of capital adequacy is attributed to liquidity (6 percent), asset
quality (4 percent), and pro?tability (18 percent) shocks in the recursive VAR. For asset
quality at same period, 22 percent of the error in the forecast is attributed to liquidity
(8 percent), capital adequacy (11 percent), and pro?tability (18 percent) shocks in the
recursive VAR. Finally, for Pro?tability at the same period, 36 percent of the error in
the forecast is attributed to liquidity (2 percent), capital adequacy (32 percent),
Figure 5.
Accumulated responses
of covariates and
their residuals
1 2 3 4 5 6 7 8 9 10
Accumulated Response of WLCR
to cholesky one S.D. innovations
–0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
WLCR WNPA
WCAR WROA
1 2 3 4 5 6 7 8 9 10
Accumulated response of RESID01
to cholesky one S.D. innovations
–0.008
–0.004
0
0.004
0.008
0.012
0.016
RESID01 RESID02
RESID03 RESID04
JFEP
6,1
38
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 6.
Residuals of the covariates
–
0
.
1
6
–
0
.
1
2
–
0
.
0
8
–
0
.
0
4 0
0
.
0
4
0
.
0
8
0
.
1
2
0
.
1
6
1 – 9 8
3 – 9 9
5 – 0 0
7 – 0 1
9 – 0 2
1 1 – 0 3
1 3 – 0 4
1 5 – 0 5
1 7 – 0 6
1 9 – 0 7
2 1 – 0 8
2 3 – 0 9
2 6 – 9 8
2 8 – 9 9
3 0 – 0 0
3 2 – 0 1
3 4 – 0 2
3 6 – 0 3
3 8 – 0 4
4 0 – 0 5
4 2 – 0 6
4 4 – 0 7
4 6 – 0 8
4 8 – 0 9
5 1 – 9 8
5 3 – 9 9
5 5 – 0 0
W
L
C
R
R
e
s
i
d
u
a
l
s
–
0
.
4
–
0
.
2 0
0
.
2
0
.
4
0
.
6
0
.
8
1 – 9 8
3 – 9 9
5 – 0 0
7 – 0 1
9 – 0 2
1 1 – 0 3
1 3 – 0 4
1 5 – 0 5
1 7 – 0 6
1 9 – 0 7
2 1 – 0 8
2 3 – 0 9
2 6 – 9 8
2 8 – 9 9
3 0 – 0 0
3 2 – 0 1
3 4 – 0 2
3 6 – 0 3
3 8 – 0 4
4 0 – 0 5
4 2 – 0 6
4 4 – 0 7
4 6 – 0 8
4 8 – 0 9
5 1 – 9 8
5 3 – 9 9
5 5 – 0 0
W
C
A
R
R
e
s
i
d
u
a
l
s
–
0
.
3
–
0
.
2
–
0
.
1 0
0
.
1
0
.
2
0
.
3
0
.
4
1 – 9 8
3 – 9 9
5 – 0 0
7 – 0 1
9 – 0 2
1 1 – 0 3
1 3 – 0 4
1 5 – 0 5
1 7 – 0 6
1 9 – 0 7
2 1 – 0 8
2 3 – 0 9
2 6 – 9 8
2 8 – 9 9
3 0 – 0 0
3 2 – 0 1
3 4 – 0 2
3 6 – 0 3
3 8 – 0 4
4 0 – 0 5
4 2 – 0 6
4 4 – 0 7
4 6 – 0 8
4 8 – 0 9
5 1 – 9 8
5 3 – 9 9
5 5 – 0 0
W
N
P
A
R
e
s
i
d
u
a
l
s
–
0
.
0
6
–
0
.
0
4
–
0
.
0
2 0
0
.
0
2
0
.
0
4
1 – 9 8
3 – 9 9
5 – 0 0
7 – 0 1
9 – 0 2
1 1 – 0 3
1 3 – 0 4
1 5 – 0 5
1 7 – 0 6
1 9 – 0 7
2 1 – 0 8
2 3 – 0 9
2 6 – 9 8
2 8 – 9 9
3 0 – 0 0
3 2 – 0 1
3 4 – 0 2
3 6 – 0 3
3 8 – 0 4
4 0 – 0 5
4 2 – 0 6
4 4 – 0 7
4 6 – 0 8
4 8 – 0 9
5 1 – 9 8
5 3 – 9 9
5 5 – 0 0
W
R
O
A
R
e
s
i
d
u
a
l
s
Banking stability
measures
39
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
and asset quality (1 percent) shocks in the recursive VAR. These observations
illustrate and underscore the long-run impact of the volatilities in the variables of
banking stability.
Forecasting
State-of-the-art VAR forecasting systems contain more than three variables and allow
for time-varying parameters to capture important drifts in coef?cients (Sims, 1993).
Multistep ahead forecasts, computed by iterating forward the recursive VAR, are
assessed in Table VI.
The ?rst two forecast error statistics largely depend on the scale of the dependent
variable and are used as relative measures to compare forecasts for the same series
across different models; the smaller the error, the better the forecasting ability of
that model according to that criterion. Very low scores of root mean squared error
(RMSE) and mean absolute error (MAE) for the forecasts indicate the strength and
accuracy of the forecast based on the VAR model. The RMSE is computed using the
formula:
RMSE ¼
?????????????????????
P
ð y 2 ^ yÞ
2
n 2k 21
s
¼
????????????????????
RSS
n 2k 21
r
Period SE WLCR WCAR WNPA WROA
Variance decomposition of WLCR
1 0.014535 100.0000 0.000000 0.000000 0.000000
4 0.026667 72.61941 21.00196 2.486288 3.892338
6 0.030266 57.23783 32.01421 7.160574 3.587378
8 0.033300 47.41230 38.09176 10.85899 3.636939
10 0.035722 41.47237 41.72264 12.37306 4.431924
Variance decomposition of WCAR
1 0.060303 21.03585 78.96415 0.000000 0.000000
4 0.110671 13.49430 74.52642 1.480222 10.49906
6 0.134815 9.291467 74.32057 3.195048 13.19292
8 0.155314 7.019326 73.10044 4.095404 15.78483
10 0.173355 5.685542 71.53319 4.259851 18.52141
Variance decomposition of WNPA
1 0.033293 10.45330 4.10 £ 10
25
89.54666 0.000000
4 0.055051 6.097703 7.514617 84.50565 1.882026
6 0.064083 6.688434 9.529185 80.91078 2.871603
8 0.068821 7.650106 10.48979 78.81341 3.046687
10 0.070980 8.136431 11.19790 77.72246 2.943210
Variance decomposition of WROA
1 0.007118 0.592220 6.053441 3.433475 89.92086
4 0.012181 4.456874 16.09711 1.998224 77.44779
6 0.014187 3.376669 23.49209 1.480018 71.65122
8 0.015846 2.709682 28.77073 1.206576 67.31302
(1.40375) (5.02008) (0.67516) (5.30725)
10 0.017332 2.269778 32.50345 1.022762 64.20401
Notes: Cholesky ordering: WLCR, WCAR, WNPA, WROA; SE – Monte Carlo (100 repetitions)
Table V.
Results of variance
decomposition analysis
JFEP
6,1
40
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The remaining two statistics are scale invariant. The Theil inequality coef?cient
always lies between 0 and 1, where 0 indicates a perfect ?t. Further, as the ultimate
test of a forecasting model is its out-of-sample performance, Table VI focuses on
pseudo out-of-sample forecasts[1] over the period 1996-2009 (Figure 7).
5. Conclusion
We provide in this study an empirical evidence for the centrality of banking system
stability for aiding ?nancial stability in the context of banking dominated emerging
economy. Employing the bank speci?c variables of banking stability namely,
liquidity, asset quality, capital adequacy and pro?tability, this study has made two
valuable contributions. First, it has analysed the banking system stability by
employing VAR technique and has established the interdependence and comovement
of the banking stability covariates to the satisfaction of economic logic. Second, this
study is unique among the evolving body of literature that underscores the
signi?cant relationship between banking system resilience and ?nancial stability.
Further, the study has enabled us to understand that the ?nancial system and more
speci?cally the banking system in India has demonstrated continued stability
compared to other countries.
A well-functioning banking system is essential to sustain economic growth, both
to prop up the economic activities in the short run and to allocate resources ef?ciently
over the longer run. Capital market capitalization also helps strengthening ?nancial
stability. The overall approach to sustain ?nancial stability has to be multi-pronged.
Ensuring overall macroeconomic balance, enhancement in the macro-prudential
functioning of institutions and markets, and reinforcement of micro-prudential
institutional soundness through regulation and supervision need to be regularly
undertaken towards ?nancial stability. Financial markets are rapidly growing by
way of technology adoption, product innovation, and geographic and sectoral
integration. This swift development of ?nancial markets while contributing to
enhanced ?nancial stability may also throw up both bene?ts and new sources of
risks to ?nancial system.
Forecast statistics WLCRF WCARF WNPAF WROAF
Root mean squared error
a
0.020435 0.095363 0.094401 0.010598
Mean absolute error
b
0.009820 0.046692 0.050487 0.005676
Mean absolute percentage error 93.65728 38.29840 300.2287 73.03435
Theil inequality coef?cient 0.108798 0.105338 0.240491 0.179451
Bias proportion 0.000000 0.027040 0.013477 0.072535
Variance proportion 0.019645 0.044505 0.044977 0.003227
Covariance proportion 0.980355 0.928455 0.941546 0.924237
Notes:
a
The mean squared forecast error is computed as the average squared value of the forecast
error over the 1996-2009 out-of-sample period, and the resulting square root is the root mean squared
forecast error reported in the table; root mean squared errors (RMSEs) are the errors squared before
they are averaged and give a relatively high weight to large errors, which infers that RMSE is most
useful when large errors are particularly undesirable;
b
mean absolute error (MAE), which is a linear
score (that all the individual differences are weighted equally in the average), measures the magnitude
of the errors in a set of forecasts without considering their direction and measures accuracy for
continuous variable; entries are the root mean square error of forecasts computed recursively for VARs
Table VI.
Forecast statistics for
the variables
Banking stability
measures
41
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 7.
Forecasting
for the covariates
–
0
.
1 0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
1 – 9 6
3 – 9 8
5 – 0 0
7 – 0 2
9 – 0 4
1 1 – 0 6
1 3 – 0 8
1 6 – 9 6
1 8 – 9 8
2 0 – 0 0
2 2 – 0 2
2 4 – 0 4
2 6 – 0 6
2 8 – 0 8
3 1 – 9 6
3 3 – 9 8
3 5 – 0 0
3 7 – 0 2
3 9 – 0 4
4 1 – 0 6
4 3 – 0 8
4 6 – 9 6
4 8 – 9 8
5 0 – 0 0
5 2 – 0 2
5 4 – 0 4
5 6 – 0 6
W
L
C
R
F
±
2
S
.
E
.
W
C
A
R
F
±
2
S
.
E
.
W
N
P
A
F
±
2
S
.
E
.
W
R
O
A
F
±
2
S
.
E
.
–
1 0 1 2 3 4
1 – 9 6
3 – 9 8
5 – 0 0
7 – 0 2
9 – 0 4
1 1 – 0 6
1 3 – 0 8
1 6 – 9 6
1 8 – 9 8
2 0 – 0 0
2 2 – 0 2
2 4 – 0 4
2 6 – 0 6
2 8 – 0 8
3 1 – 9 6
3 3 – 9 8
3 5 – 0 0
3 7 – 0 2
3 9 – 0 4
4 1 – 0 6
4 3 – 0 8
4 6 – 9 6
4 8 – 9 8
5 0 – 0 0
5 2 – 0 2
5 4 – 0 4
5 6 – 0 6
–
0
.
5 0
0
.
5
1
.
0
1
.
5
2
.
0
2
.
5
1 – 9 6
3 – 9 8
5 – 0 0
7 – 0 2
9 – 0 4
1 1 – 0 6
1 3 – 0 8
1 6 – 9 6
1 8 – 9 8
2 0 – 0 0
2 2 – 0 2
2 4 – 0 4
2 6 – 0 6
2 8 – 0 8
3 1 – 9 6
3 3 – 9 8
3 5 – 0 0
3 7 – 0 2
3 9 – 0 4
4 1 – 0 6
4 3 – 0 8
4 6 – 9 6
4 8 – 9 8
5 0 – 0 0
5 2 – 0 2
5 4 – 0 4
5 6 – 0 6
–
0
.
1
0
–
0
.
0
5
0
.
0
0
0
.
0
5
0
.
1
0
0
.
1
5
0
.
2
0
0
.
2
5
1 – 9 6
3 – 9 8
5 – 0 0
7 – 0 2
9 – 0 4
1 1 – 0 6
1 3 – 0 8
1 6 – 9 6
1 8 – 9 8
2 0 – 0 0
2 2 – 0 2
2 4 – 0 4
2 6 – 0 6
2 8 – 0 8
3 1 – 9 6
3 3 – 9 8
3 5 – 0 0
3 7 – 0 2
3 9 – 0 4
4 1 – 0 6
4 3 – 0 8
4 6 – 9 6
4 8 – 9 8
5 0 – 0 0
5 2 – 0 2
5 4 – 0 4
5 6 – 0 6
JFEP
6,1
42
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Note
1. Pseudo out-of-sample forecasts are often referred to as pseudo or “simulated” out-of-sample
forecasts to emphasise that they simulate how these forecasts would have been computed in
real time, although of course this exercise is conducted retrospectively, not in real time.
References
Aka, B.E. (2006), “On the duration of the ?nancial system stability under liberalization”,
Emerging Markets Review, Vol. 7 No. 2, pp. 147-161.
Allen, W.A. and Wood, G. (2006), “De?ning and achieving ?nancial stability”, Journal of
Financial Stability, Vol. 2 No. 2, pp. 152-172.
Anzoategui, D., Soledad, M., Per? ´a, M. and Melecky, M. (2012), “Bank competition in Russia: an
examination at different levels of aggregation”, Emerging Markets Review, Vol. 13, pp. 42-57.
Beck, T., Cull, R. and Jerome, A.T. (2005), “Bank privatization and performance: evidence from
Nigeria”, Journal of Banking & Finance, Vol. 29, pp. 2355-2379.
Beck, T., Hesse, H., Kick, T. and Von Westernhagen, N. (2009), “Bank ownership and stability:
evidence from Germany”, Bundesbank Working Paper Series.
Blavarg, M. and Nimander, P. (2002), “Interbank exposures and systemic risk”, Proceedings of
the Third Joint Central Bank Research Conference on Risk Measurement and Systemic Risk,
Bank for International Settlements, Basel, pp. 287-305.
Borio, C. (2003), “Towards a macro prudential framework for ?nancial supervision and
regulation?”, BIS Working Papers No. 128, Basel.
Borio, C. and Lowe, P. (2002), “Asset prices, ?nancial and monetary stability: exploring the
nexus”, BIS Working Papers No. 114, Basel.
Boyd, J.H. and de Nicolo, G. (2005), “The theory of bank risk-taking and competition revisited”,
Journal of Finance, Vol. 60 No. 3, pp. 1329-1343.
Boyd, J.H. and Runkle, D.E. (1993), “Size and performance of banking ?rms: testing the
predictions of theory”, Journal of Monetary Economics, Vol. 31, pp. 47-67.
Calomiris, C.W. and Mason, J.R. (1997), “Contagion and bank failures during the great
depression: the June 1932 Chicago banking panic”, American Economic Review, Vol. 87
No. 5, pp. 863-883.
Calomiris, C.W. and Mason, J.R. (2000), “Causes of bank distress during the great depression”,
National Bureau of Economic Research Working Paper No. 7919, Cambridge, MA.
Crockett, A. (1997), “Maintaining ?nancial stability in a global economy”, paper presented at the
Federal Reserve Bank of Kansas City’s Symposium, Jackson Hole, WY, 28-30 August.
Davis, E.P. (1993), “Bank credit risk”, Working Paper Series No. 8, Vol. 8, Bank of England, London.
Davis, E.P. (2003), “Towards a typology for systemic ?nancial instability”, Public Policy
Discussion Papers No. 03-20, Economics and Finance Section, School of Social Sciences,
Brunel University, Uxbridge.
De Bandt, O. and Hartmann, P. (2002), “Systemic risk in banking: a survey”, in Goodhart, C. and
Illing, G. (Eds), Financial Crises, Contagion, and the Lender of Last Resort – A Reader,
Oxford University Press, London, pp. 249-298.
de Nicolo, G. (2000), “Size, charter value, and risk in banking: an international perspective”,
International Finance Discussion Papers No. 689, Board of Governors of the Federal
Reserve System, Washington, DC.
Diamond, D.W. (1984), “Financial intermediation and delegated monitoring”, Review of Economic
Studies, Vol. 51, pp. 393-414.
Banking stability
measures
43
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Diamond, D.W. and Dybvig, P. (1983), “Bank runs, deposit insurance and liquidity”, Journal of
Political Economy, Vol. 91, pp. 401-419.
Diamond, D.W. and Rajan, R. (2001), “Liquidity risk, liquidity creation, and ?nancial fragility:
a theory of banking”, Journal of Political Economy, Vol. 109, pp. 287-327.
Foot, M. (2003), “Protecting ?nancial stability – how good are we at it?”, speech given at
University of Birmingham, 6 June.
Gadanecz, B. and Jayaram, K. (2009), “Measures of ?nancial stability – a review”, in Bank for
International Settlements (Ed.), Proceedings of the IFC Conference on “Measuring
Financial Innovation and Its Impact”, Basel, 26-27 August 2008, IFC Bulletins Chapters,
Vol. 31, Bank for International Settlements, Basel, pp. 365-380.
Gauthier, C., Graham, C. and Liu, Y. (2004), “Financial conditions indexes for Canada”, Bank of
Canada Working Paper No. 2004-22.
Goodhart, C.A.E. (1987), “Why do banks need a central bank”, Oxford Economic Papers, Vol. 39,
pp. 75-89.
Goodhart, C.A.E. and Hofmann, B. (2001), “Asset prices, ?nancial conditions, and the
transmission of monetary policy”, paper presented for the Conference on Asset Prices,
Exchange Rates, and Monetary Policy, Stanford University, 2-3 March.
Goodhart, C.A.E., Sunirand, P. and Tsomocos, D.P. (2004), “A model to analyse ?nancial fragility:
applications”, Journal of Financial Stability, Vol. 1, pp. 1-35.
Goodhart, C.A.E., Sunirand, P. and Tsomocos, D.P. (2005), “A risk assessment model for banks”,
Annals of Finance, Vol. 1, pp. 197-224.
Goodhart, C.A.E., Sunirand, P. and Tsomocos, D.P. (2006a), “A model to analyse ?nancial
fragility”, Economic Theory, Vol. 27, pp. 107-142.
Goodhart, C.A.E., Sunirand, P. and Tsomocos, D.P. (2006b), “A time series analysis of ?nancial
fragility in the UK banking system”, Annals of Finance, Vol. 2, pp. 1-21.
Goodhart, C.A.E., Hartmann, P., Llewellyn, D., Rojas-Suarez, L. and Weisbrod, S. (1998),
Financial Regulation: Why, How and Where Now?, Routledge, London.
Gray, D.F., Merton, R.C. and Bodie, Z. (2007), “New framework for measuring and managing
macro-?nancial risk and ?nancial stability”, NBER Working Paper No. 13607.
Haldane, A., Hall, S., Saporta, V. and Tanaka, M. (2004), “Financial stability and macroeconomic
models”, Bank of England Financial Stability Review, June, pp. 80-88.
Hasan, I. and Dwyer, G. (1994), “Bank runs in the free banking period”, Journal of Money, Credit,
and Banking, Vol. 26, pp. 271-288.
Hawkins, J. and Klau, M. (2000), “Measuring potential vulnerabilities in emerging market
economies”, BIS Working Papers No. 91, October, Vol. 91.
Hellwig, M. (1991), “Banking, ?nancial intermediation, and corporate ?nance”, in Giovannini, A.
and Mayer, C. (Eds), European Financial Integration, Cambridge University Press,
Cambridge, pp. 35-63.
Hoggarth, G.A., Milne, A. and Wood, G.E. (1998), “Alternative routes to banking stability:
a comparison of UK and German banking systems”, Financial Stability Review, Vol. 5,
pp. 55-68.
Houben, A., Kakes, J. and Schinasi, G. (2004), “Toward a framework for safeguarding ?nancial
stability”, IMF Working Papers No. WP/04/101, International Capital Markets
Department, International Monetary Fund, Washington, DC.
IMF (2006), Financial Soundness Indicators: Compilation Guide, International Monetary Fund,
Washington, DC, March.
JFEP
6,1
44
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
IMF (2008), Global Financial Stability Report, International Monetary Fund, Washington, DC, April.
Issing, O. (2003), “Monetary and ?nancial stability: is there a trade-off?”, paper presented at the
Conference on Monetary Stability, Financial Stability and the Business Cycle, Bank for
International Settlements, Basel, 28-29 March.
Merton, R. (1974), “On the pricing of corporate debt: the risk structure of interest rates”, Journal
of Finance, Vol. 29, pp. 449-470.
Mishkin, F.S. (1994), “Global ?nancial instability: framework, events, issues”, Journal of
Economic Perspectives, Vol. 13, pp. 3-25.
Mishra, R.N., Majumdar, S. and Bhandia, D. (2013), “Banking stability – a precursor to ?nancial
stability”, Reserve Bank of India Working Paper Series No. WPS (DEPR): 01/2013, January.
Montagnoli, A. and Napolitano, O. (2004), “Financial condition index and interest rate settings:
a comparative analysis”, Money Macro and Finance Research Group Conference Series No. 1.
Morgan, J.P. (1999), CreditMetrics, 4th ed., available at: www.creditmetrics.com
Nelson, W.R. and Perli, R. (2005), “Selected indicators of ?nancial stability”, paper presented at
the 4th Joint Central Bank Research Conference on “Risk Measurement and Systemic
Risk”, ECB, Frankfurt am Main, November.
Pain, D. (2003), “The provisioning experience of the major UK banks: a small panel
investigation”, Working Paper No. 177, Bank of England, London.
Saunders, A. and Wilson, B. (1996), “Contagious bank runs: evidence from the 1929-33 period”,
Journal of Financial Intermediation, Vol. 5 No. 4, pp. 409-423.
Schinasi, G.J. (2004), “De?ning ?nancial stability”, IMF Working Paper No. WP/04/187,
International Capital Markets Department, International Monetary Fund, Washington, DC.
Sims, C. (1980), “Macroeconomics and reality”, Econometrica, Vol. 48 No. 1, pp. 1-48.
Tsomocos, D.P. (2003a), “Equilibrium analysis, banking, and ?nancial instability”, Journal of
Mathematical Economics, Vol. 39, pp. 619-655.
Tsomocos, D.P. (2003b), “Equilibrium analysis, banking, contagion and ?nancial fragility”,
Bank of England Working Paper Series No. 175.
Further reading
Bernanke, B. (2005), “The global saving glut and the US current account de?cit”, Homer Jones
Lecture, St Louis, MO, 14 April, available at: www.federalreserve.gov
Grossman, R. (1993), “The macroeconomic consequences of bank failures under the national
banking system”, Explorations in Economic History, Vol. 30, pp. 294-320.
Group of Ten (2001), Report on Consolidation in the Financial Sector, BIS, IMF, OECD, Basel.
Laeven, L. and Valencia, F. (2008), “Systemic banking crises: a new database”, IMF Working
Paper No. WP/08/224, International Monetary Fund, Washington, DC.
About the author
Dr Vighneswara Swamy is currently with IBS-Hyderabad as an Associate Professor in the
?nance area. His research interests include ?nancial stability, ?nancial intermediation, risk
management, ?nancial markets, banking and ?nance and development economics, etc. He has
published several papers in journals of international recognition as well as national-level
publications. Vighneswara Swamy can be contacted at: [email protected]
Banking stability
measures
45
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_816830893.pdf