Financial and economic stability as two sides of a coin

Description
The purpose of this paper is to provide a different context for considering issues of financial
stability and instability, with reference to economic growth and price stability in particular

Journal of Financial Economic Policy
Financial and economic stability as ‘two sides of a coin’: Non-crisis regime
evidence from the UK based on VECM
Muhammad Ali Nasir Mushtaq Ahmad Ferhan Ahmad J unjie Wu
Article information:
To cite this document:
Muhammad Ali Nasir Mushtaq Ahmad Ferhan Ahmad J unjie Wu , (2015),"Financial and economic
stability as ‘two sides of a coin’", J ournal of Financial Economic Policy, Vol. 7 Iss 4 pp. 327 - 353
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-01-2015-0006
Downloaded on: 24 January 2016, At: 21:53 (PT)
References: this document contains references to 60 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 48 times since 2015*
Users who downloaded this article also downloaded:
Md. Dulal Miah, Kashfia Sharmeen, (2015),"Relationship between capital, risk and efficiency:
A comparative study between Islamic and conventional banks of Bangladesh", International
J ournal of Islamic and Middle Eastern Finance and Management, Vol. 8 Iss 2 pp. 203-221 http://
dx.doi.org/10.1108/IMEFM-03-2014-0027
Mohamad Sadegh Sangari, Reza Hosnavi, Mohammad Reza Zahedi, (2015),"The impact of
knowledge management processes on supply chain performance: An empirical study", The
International J ournal of Logistics Management, Vol. 26 Iss 3 pp. 603-626http://dx.doi.org/10.1108/
IJ LM-09-2012-0100
Louise Fletcher, Yiannis E. Polychronakis, (2007),"Capturing knowledge management
in the supply chain", EuroMed J ournal of Business, Vol. 2 Iss 2 pp. 191-207 http://
dx.doi.org/10.1108/14502190710826059
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.
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
*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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Financial and economic stability
as ‘two sides of a coin’
Non-crisis regime evidence from the UK
based on VECM
Muhammad Ali Nasir
Faculty of Business and Law, Leeds Beckett University, Leeds, UK
Mushtaq Ahmad
Management Sciences Department, Comsats Institute of IT,
Wah Cantt, Pakistan
Ferhan Ahmad
Graduate School of Business, Faculty of Business and Economics,
Monash University, Melbourne, Australia, and
Junjie Wu
Faculty of Business and Law, Leeds Beckett University, Leeds, UK
Abstract
Purpose – The purpose of this paper is to provide a different context for considering issues of fnancial
stability and instability, with reference to economic growth and price stability in particular.
Design/methodology/approach – This paper pursued an empirical exploration of six pillars of
fnancial stability, based on a data set for the UKextending from1985 (Q1) to 2008 (Q2), through the
construction of a vector error correction model, including an impulse response function analysis.
Findings – The fndings show a strong association between the fnancial and economic stability
even in a non-crisis regime. This includes, for example, a strong association exists between the
stock market and the real economy; exchange rate appreciation may not provide for long-term real
economic growth; infation does not contribute to real economic growth, both the sensitivity of the
economy to yields and a signifcant lag in transitional effects from fnancial markets to the real
sector; a positive role of credit creation within a non-crisis regime; exchange rate appreciation
affects purchasing power; and potential points of linkage between sovereign debt activity and
general price levels.
Research limitations/implications – The fndings should be considered in the context of a concept
of the economy as fundamentally dynamic and subject to complex cumulative processes.
Practical implications – The fndings indicate there is a role for state oversight and intervention
within a non-crisis regime based on the complexity of possible interactions that may undermine
fnancial and price stability, with consequences for their association with economic growth.
JEL classifcation – B26, E44, G01
The authors acknowledge and thank Dr Jamie Morgan for his constructive comments which led
to signifcant improvement of the original draft. They also appreciate comments received at
BMRC-DEMS Conference on Macro and Financial Economics/Econometrics held in May 2014 at
Brunel University, UK, and International Academy of Business and Economics Conference held in
July 2014 at University of Verona, Italy.
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1757-6385.htm
Financial and
economic
stability
327
Received18 January2015
Revised11 March2015
26 May2015
Accepted26 May2015
Journal of Financial Economic
Policy
Vol. 7 No. 4, 2015
pp. 327-353
©Emerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-01-2015-0006
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Originality/value – The study provides a newperspective for considering issues of fnancial stability
and instability.
Keywords Financial markets, Economic development: fnancial markets,
Financial markets and the macroeconomy, Financial economics, Foreign exchange
Paper type Research paper
1. Introduction
The Global Financial Crisis, which began September 2008, resulted in a fundamental
shift in perception regarding the fnancial sector. Borio (2011, p. 26) argues that
“fnancial and macroeconomic stabilities are two sides of the same coin”. However, an
important issue is how to defne the stability of the fnancial sector. Foot (2003) notes
there is currently no single and comprehensive defnition of fnancial stability and that
it is diffcult to derive one. However, he argues it could be defned in the context of
fnancial asset price volatility and the generality of fnancial markets and institutions.
Concomitantly, Goodhart (2004, p. 2) also notes that “Indeed there is currently no good
way to defne” fnancial stability. Khorasgani (2010, pp. 20-21) states that “There is no
consensus on a defnition of fnancial instability”. Khorasgani’s study suggests a need
for a broader conception. To provide such a conception, Filardo (2008) suggests
beginning from better measures of fnancial instability. For the purpose of this study,
fnancial instability is defned as a situation in which economic performance is
signifcantly impaired by fuctuations in the price of fnancial assets, or in the ability of
fnancial intermediaries to meet their contractual obligations.
A signifcant issue is how to appropriately measure and contextualize fnancial
instability in relation to fnancial stability. Seminal work has been done by Reinhart and
Rogoff (2009)[1] based on six fnancial aggregates, each associated with a specifc
threshold in terms of which the onset of fnancial instability is defned: a currency crisis
(a Forex decline of 15 per cent.), an infation crisis (20 per cent Consumer Price Index
[CPI]), a stock market crisis (an index decline of 25 per cent), a foreign debt crisis
(government default), a domestic debt crisis (default) and a banking crisis (a run,
bankruptcy, etc.). However, one potential problem with this approach is that the
threshold itself becomes a focus of analysis (see some of the problems identifed by Ho,
2004; Hagen and Ho, 2007)[2]. More specifcally, any state of affairs belowthe threshold
is by implication deemed stable.
This paper adopts a more inclusive approach to aggregates by placing a broader
context around the issue of fnancial stability and instability. Rather than seek to
identify a threshold. It explores the implications of fnancial aggregates within broader
economic contexts and over a duration that is not restricted to one designated as a period
of crisis. Specifcally, UK data from the period 1985 (Q1) to 2008 (Q2)[3] are used. In so
doing, a different set of aggregates than those chosen by Reinhart and Rogoff (2009) is
selected. It terms these six pillars of fnancial stability: the bond market yield, the
domestic lending rate to UK residents, the infation rate, the spot exchange rate, GDP,
the defcit to GDP ratio and stock-market returns based on an index measure. The
purpose of taking this approach is to constructively move beyond a narrow focus on
periods of crisis.
Much of the research conducted on fnancial instability has focused on extreme
disturbances in the fnancial sector (Bordo et al., 2003).
JFEP
7,4
328
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
However, fnancial instability is not necessarily the same as crisis. If one does not
focus on thresholds, one can also capture periods of relative instability in relation to
stability. This can allow a broader analysis of the relation between aspects of stability
and instability. This is important because it facilitates a focus on fnancial stability that
does not neglect the lesser episodes or instances of instability. These too are signifcant
for an adequate understanding of an economy. This is so in two ways:
(1) As a matter of concern for public policy, periods of relative fnancial stability are
also signifcant in terms of economic performance. The focus should not simply
be on what is to be avoided, but also on what conditions are associated with more
stability.
(2) As a corollary of 1, in terms of focus, the ex ante tendency towards an economic
disturbance, not the actual damage ex post, is equally as signifcant in terms of
instability as the crossing of a given threshold.
The paper proceeds as follows, Section 2 briefy reviews the existing evidence on the
importance of fnancial stability for economic stability. Section 3 sets out and justifes
the defnitions and data sources for the six pillars of fnancial stability. Section 4 sets a
vector error correction model (VECM) as a means to analyze the association between the
stated six pillars of fnancial stability and economic stability/instability and then
analyze the fndings. The objective of this paper is to test the possible associations of
these six pillars with economic stability (represented by real GDP) and then, given the
historical/theoretical emphasis placed upon it, price stability (represented by the CPI).
As such, the presentation of the model to a reduced two-equation form is restricted. It
augments the model fndings with an impulse response function (IRF). The key point to
emphasize is that important associations can be identifed between economic stability,
fnancial stability and price stability based on the associations that arise through the six
pillars of fnancial stability. It then follows, based on the IRF, that instability creates
adverse effects that it would be argued providing a justifcation for macro-prudential
approaches to the fnance system.
2. The signifcance of fnancial (in)stability and the limits of price
stability
As noted in the introduction, there is a general lack of consensus regarding a
comprehensive defnition of both fnancial instability and stability. However, for the
purposes of this study, it requires only a generic working defnition of each, as the basis
of the approach resists a focus on thresholds and so also initially requires no defnitive
translation of the general concept of (in)stability into specifc quantities. As such it
adopts Schioppa’s (2002, p. 20) working defnition:
[…] fnancial stability is a condition where the fnancial system is able to withstand shocks
without giving way to cumulative processes, which impair the allocation of savings to
investment opportunities and the processing of payments in the economy[4].
Concomitantly, fnancial instability may be generally defned as a situation in which
economic performance is potentially impaired by fuctuations in the price of fnancial
assets, or in the ability of fnancial intermediaries to meet their contractual
obligations[5].
329
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
The signifcance of the fnance sector and of fnancial stability for the real economy
has been recognized for more than a century (for seminal work see Bagehot, 1873). The
prominence of analyses of the adverse consequences of fnancial instability tends to
follow on from the manifestation of fnancial crises (Minsky, 1974, 1982; Kindleberger,
1978), and there have been many different approaches to the broad conceptual problem
of investment, saving and the role of banking (some of which currently revolve around
issues of endogenous vs exogenous money). Most of the existing studies of fnancial
stability have been from the perspective of the causes and consequences of instability,
subject to a focus on extreme and relatively rare events. Conversely, a great deal of
economics theory is about the forces that create market equilibrium rather than
addressing those that create market disequilibrium, ignoring the kinds of endogenous
behavioral processes explored by Minsky or Kindleberger within fnance. It is only
relatively recently that the burgeoning fnance literature has started to consider the two
in a more nuanced way (Gertler, 1988).
One major issue to arise from the recent global fnancial crisis has been an
acknowledgment of the problem of price stability. It has been widely recognized that
although price stability is a desirable aspect of an economy, the existence of such
price stability is not in itself a guarantor of long-termfnancial stability. However, price
stability and fnancial stability are important contributory factors for economic
stability, as they create sustainable confdence among depositors and investors,
resulting in stabilizing expectation effects expressed, for example, in bank relations and
through money markets (Arouri et al., 2013). Moreover, one of the permissive causes of
the recent fnancial crisis was the failure to recognize the need for both price and
fnancial stability to be focused upon, creating a policy blind spot in terms of warnings
regarding macroeconomic stability (leading then to policy calls for macro-prudential
regulation; for issues see Borio, 2011; Blanchard et al., 2010; Gros, 2010; White, 2006)[6].
In this context, Borio states: “fnancial and macroeconomic stabilities are two sides of
the same coin” (Borio, 2011, p. 26). Given this, it seems reasonable to develop a focus on
the relations between aspects of fnance and the macro-economy and not to restrict such
a focus to periods of manifest fnancial instability, but rather to explore also periods of
apparent stability to provide insights into those relations.
Intuitively, one would anticipate that there are relations between fnancial and
macro-economic stability, including price stability. This is contingently confrmed by
Bordo et al. (2003), who explore the impact of price levels on fnancial stability in the UK
based on data from1796-1999. According to Bordo et al. (2003), price level shocks had a
considerable impact on fnancial instability in the UK[7]. However, the Bordo study is
uni-dimensional and was only able to identify the impact of price (in)stability on
fnancial (in)stability. This limits its insight for policy. This study seeks to move beyond
this (see Sections 3 and 4 hereafter). In doing so, however, it is recognized that policy
involves crisis and non-crisis regimes (Nasir and Soliman, 2014). For example, Martin
and Milas (2012) fnd (using monthly data, M) that there have been signifcant
differences in monetary policy in the UKbetween periods of crisis (2007M5-2010M7) and
non-crises (1992M10-2007M4)[8]. Baxa et al. (2013) also note such differences across a
range of major central banks (The Fed, Bank of England, Reserve Bank of Australia,
Bank of Canada and Sveriges Riskbank), though the scale of intervention varied[9].
According to Baxa et al. (2013), the central bank focus during periods of stress was
primarily on the stock market and banking. However, this paper follows Ostry et al.
JFEP
7,4
330
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
(2012) and proposes that foreign exchange rates are also a signifcant area of concern in
terms of fnancial stability[10].
In the broadest terms, the fnancial sector consists of several components including
capital markets, sovereign debt markets and foreign exchange markets. As there are
different components, there can also be different aspects of (in)stability related to these
components (Filardo, 2008). As noted in the introduction, Reinhart and Rogoff (2009)
explore these components based on six fnancial aggregates as a measure of fnancial
instability. This study takes a slightly different approach, focusing on six fnancial
aggregates as “pillars of fnancial stability” and does so without a focus on thresholds or
a default to periods of manifest crisis. It sets these out in the next section.
3. Data and the six pillars of fnancial stability
This study aims to analyze how measures of fnancial stability perform in periods of
relative stability. It uses UK data from the period 1985 (Q1) to 2008 (Q2). The data are
obtained from the Bank of England, from the World Bank’s database World
Development Indicators, from the FTSE 100 historical prices archive and from the
Offce of National Statistics. It stops at Q2 2008 because the global fnancial crisis
manifests in Q3 and this period is already well-covered by a variety of other researchers
and with reference to Reinhart and Rogoff’s thresholds. In other word, its focus is on
non-crisis.
It considers the following seven fnancial aggregates as indicators of fnancial and
economic stability:
(1) Bond market yield (BMY hereafter): It is the reconstructed raw monthly real
government liability curve data from the Bank of England UK yield curve
(forward curve) on 10-year UKGovernment bonds (Gilts) as a quarterly average,
1985 (Q1) to 2008 (Q3)[11]. This data set (following Campbell, 1995) is selected
because it is adjusted for infation and because the forward yield curve
incorporates the expectations of signifcant economic agents and so also refects
the decision-making process of those agents that may manifest as forms of
observable stability and instability (fuctuations). As such it refects the
confdence of market participants and investors in bonds as well as returns on
investment (Gulley and Sultan, 2003; Nasir and Soliman, 2014). The yield on
bonds is also important for the government, as it represents its borrowing cost.
It is considered this cost to be of interest at any level rather than, following
Reinhart and Rogoff (2009), an indicator that signals an ostensibly (which is now
disputed, see Hernden et al., 2014) prohibitive threshold.
(2) Domestic lending to income ratio for UK residents (DL hereafter): A linear
interpolation to the World Bank annual domestic credit to private sector (% of
GDP) data for the UK to derive quarterly estimates, 1985 (Q1) to 2008 (Q3), is
applied[12]. This data set is selected because it encompasses a wide range of
lending organizations and also both secured and unsecured lending[13]. As such
it provides a relatively comprehensive picture of lending. If one acknowledges
the interconnection between borrowing, consumption and investment, then
changes in the ratio of domestic lending to income are of signifcance both as
sources of economic activity and of potential future (in)stability prior to any
other manifestation of crisis[14]. As noted in the introduction, it may have ex
ante signifcance.
331
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
(3) Stock-market returns (SMR hereafter): The raw monthly adjusted closing price
data for the FTSE 100 index to a quarterly average, 1985 (Q1) to 2008 (Q3), are
converted[15]. This data set is selected rather than the FTSE All Share, 350 or
250 because the 100 constitutes approximately 80 per cent of the capitalization of
equity markets and contains most of the corporations which institutional
investors are empowered to trade and so is the main focus of signifcant trading
activity. Stock-market activity is widely acknowledged to be important because
it creates wealth effects (Airaudo et al., 2008, and Tsouma, 2009), and is also a
potential site where adverse effects can begin or be observed (Friedman, 1998,
for example, notes that a rise in stock prices may create a knock-on effect to
rising expected/offered returns on risky assets, while a rise in the volume and
frequency of fnancial transactions creates a further transactional demand for
money; see also Funke et al., 2011). Again, as with the BMY, SMR to be of
potential interest at levels below any threshold set to capture manifest crisis is
considered, recalling that Reinhart and Rogoff (2009) identify an index decline of
25 per cent as a threshold, which for an index rather than a given sector or equity
is both relatively rare and more indicative of ex post damage.
(4) Infation rate (INF hereafter): Offce of National Statistics monthly Retail Price
Index (RPI) series data (base year 1974) to a quarterly average, 1985 (Q1) to 2008
(Q3), are converted[16]. Although the Treasury and Bank of England’s preferred
measure of infation is now the CPI, it has selected the RPI because it was the
preferred measure for much of the period under analysis (and the CPI was only
introduced in 1992). Moreover, the RPI is more comprehensive, as it includes
items such as mortgage payments, rents and council tax (and is based on a
different calculation of the mean – arithmetic rather than geometric); the RPI is
typically higher than the CPI[17]. Since the 1980s, price stability has been a
primary concern of modern monetary policy pursued through central banks
(focused initially on the money supply but later on nominal interest rates), and
the framework of concerns of both modern economic theory and general
economic policy has also supported this (based on a policy ineffectiveness
approach to tradeoffs between infation and a natural rate of unemployment
unless supply side issues are addressed). Concerns over price stability arise far
in advance of Reinhart and Rogoff’s threshold of 20 per cent CPI; the Bank of
England, for example, target 2 per cent.
(5) Spot exchange rate (SER hereafter): The Bank of England quarterly average
spot exchange rate for Sterling against the US$, 1985 (Q1) to 2008 (Q3), is
used[18]. This data set is selected because the US$ is the unoffcial reserve
currency of the international fnancial system and represents the most
signifcant traded currency on international markets for the period under
analysis (despite that there has also been diversifcation into holdings in other
currencies, including new currencies, such as the euro). Spot rates are those
observed (to be executed two days hence) by the Bank of England’s Foreign
Exchange Desk in the London interbank market at 4 p.m. each day[1920]. They
provide an authoritative but not offcial guide to the exchange rate and its
movements. The exchange rate is widely recognized to be of potential
signifcance in terms of (in)stability because movements can signal capital fight
JFEP
7,4
332
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
and repatriation but also because stable exchange rates are a desirable condition
for trade and investment (Malikane and Semmler, 2008; Khorasgani, 2010).
(6) Defcit to GDP ratio (DGDPR hereafter): The Offce of National Statistics
quarterly data for public sector current budget defcit/surplus as a per cent of
GDP (rolling four quarter average), 1985 (Q1) to 2008 (Q3), are used. 20
Following Muscatelli et al. (2004), this data set as a defcit to income ratio is
selected, rather than selecting a debt to income ratio, because it provides an
on-going snapshot of the prevailing shortfall or surplus in government fnances,
rather than a cumulative expression of past debt. As such, it provides an
indication of the potential for proximate fnancial distress or its absence in recent
events. This then is signifcant for issues of (in)stability.
(7) Gross domestic product growth rate (GDPG hereafter): The Offce of National
Statistics quarterly data for GDP at current market prices, 1985 (Q1) to 2008
(Q3), are used[21]. This data set is selected because it is adjusted for infation and
represents real economic growth. GDP is a primary point of reference for all
other indicators of economic and fnancial stability in terms of the business
cycle. In what follows its main signifcance is as a dependent variable.
4. Econometric model and data analysis
The objective of this study is to explore aspects of “six-pillars” of fnancial stability in
the UK, based on a data set for the period 1985 (Q1) to 2008 (Q3). The data set includes
multiple variables and a time series. Vector autoregressive (VAR) models are widely
used for such data sets (Basu and Michailidis, 2013). VAR is used with interrelated time
series and for analyzing the dynamic impact of random disturbances on the system of
variables. In a VAR model, the endogenous and explanatory variables interact
simultaneously, hence there is an extended information set, which makes it a more
adequate presentation of key aspects of an economic system than a standard multiple
regression model (Pecican, 2010). The VAR approach sidesteps the need for structural
modeling by modeling every endogenous variable in the system as a function of the
lagged values of all of the endogenous variables in the system. As only lagged values of
the endogenous variables appear on the right-hand side of each equation, there is no
issue of simultaneity. Importantly, the assumption that any disturbances are not serially
correlated is not restrictive, because any serial correlation could be absorbed by adding
more lagged y’s. As such, using VAR, any serial correlation of errors does not become an
issue.
Having chosen a VAR approach, to establish whether the data are stationary, frst a
unit root test (augmented Dickey and Fuller [ADF] test; Section 4.1) is applied.
Stationarity is important to establish because in its absence, a drift in the data series
would yield potentially spurious results, undermining the model and placing any
association between the given variables in question. One is then required to distinguish
whether an unstructured or simple VARor a restricted or VECMis appropriate (Section
4.2). VECMis the appropriate model when co-integration exists among the variables. As
such, a Johansen co-integration test to establish that co-integration pertains (Section 4.3)
is performed. Based on the fndings, then a VECMwas run. As the objective is to test the
possible associations of the six pillars of fnancial stability with economic stability
(represented by real GDP) and then, given the historical/theoretical emphasis placed
upon it, price stability (represented by the RPI), a reduced form of the full VECM is
333
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
presented, which specifes only two equations, based on the foci (Section 4.4). These
results are then separately explored (Sections 4.5 and 4.6). The estimations of
parameters are then tested for robustness using three standard diagnostic tests:
heteroskedasticity (White), autocorrelation (Breusch-Godfrey serial correlation LM) and
exogeneity (Wald) (Sections 4.51 and 4.61).
The empirical results obtained from the VECM provide some insight regarding
economic stability but also involve an inherent limitation[22]. Growth may respond
differently to the dynamics of any given variable, e.g. equity prices, from period to
period. If one simply estimates parameters within the VEC for each of the lagged
periods, then one will obtain a range of coeffcients (different for each quarter;
potentially ranging in terms of sign, size and signifcance). Here, the results are diffcult
to interpret in terms of an overall responsiveness of the dependent variable to the
explanatory variables. An IRF analysis allows us to address this issue (Canova, 2007, p.
130; Sections 4.52 and 4.62).
The IRFs are obtained from the moving average representation of the original VAR
model. The IRFs are the dynamic response of each endogenous variable to a one-period
standard deviation shock to the system. The responsiveness of the dependent variables
in the VAR to shocks on each variable is revealed by the impulse responses. So, for each
variable from each equation separately, a unit shock is applied to the error, and the
effects upon the VARsystemover time are noted (Brooks, 2008). In other words, a shock
to the ith variable directly affects the ith variable and, in addition, is transmitted to the
other endogenous variables through the dynamic lag structure of the VAR. In this sense,
an IRF traces out the effects of a onetime shock to one of the innovations of the current
and future values of the endogenous variables.
4.1 Unit root test
In the event that a variable has unit roots, frst and subsequent differencing renders
them stationary (Greene, 2012). The common practice recommended by Engle and
Granger (1987) is the use of an ADF test for unit roots which is applied to the data set,
and the results are presented in Table I.
As indicated in Table I, after taking the frst difference, the results were greater than
the critical values at a 5 per cent signifcance level, which implies that all the data series
were frst difference stationary or I(1) variables. As such, stationarity was established.
4.2 The VAR model
As it has selected a VAR approach, no major methodological issues for estimation arise
regarding the order in which variables are presented (Brooks, 2008). The next procedure
is to establish the most appropriate number of lags for the explanatory variables within
the data set. To do so, an optimal lag selection test using a range of standard criteria is
performed, rather than one only. These are set out in Table II.
As one can see, the SC and HQ criteria indicate one as an optimal lag, while the LR
and FPE criteria indicate fve as the optimal lag, and the AIC criterion indicates six as
the optimal lag. For these purposes, the AIC criterion and, so, six as the optimal lag are
chosen. Here, it follows Liew (2004). The data set is for quarters and constitutes 95
observations and this is relatively small. Liewmakes the case that AICis appropriate for
small samples. Moreover, AIC minimizes the possibility of underestimation of the
optimal lag length while improving the potential that the true lag length is recovered.
JFEP
7,4
334
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
4.3 Johansen co-integration test
To distinguish whether an unstructured or simple VAR or a restricted or VECM is
appropriate, the Johansen co-integration test is used. This method is used rather than the
Engle-Granger test because unlike the Engle-Granger test, it allows more than one
co-integrating relationship (Greene, 2012). As there are six variables, there could be
more than one such relationship. The results of the Johansen test are presented in
Table III.
As Table III indicates, both of the unrestricted co-integration rank tests (trace and
max eigen statistics) showthat the null of no co-integration was rejected at the 5 per cent
level of signifcance on the basis of MacKinnon-Haug-Michelis (1999) P-values. As such,
the results of the co-integration test show that there is a co-integrating relationship
Table I.
Augmented Dickey–
Fuller unit root test
Variable ADF test stat* 1% level** 5% level*** P-value
At level I(o)
Bond yield (0.264) (3.502) (2.892) 0.925
Domestic lending (0.442) (3.502) (2.893) 0.983
Spot exchange rate (3.026) (3.502) (2.892) 0.036
GDP growth rate (2.004) (3.502) (2.892) 0.284
Defcit to GDP ratio (3.765) (3.502) (2.892) 0.004
Stock market (1.308) (3.502) (2.892) 0.622
Infation (0.294) (3.502) (2.894) 0.920
1
st
difference I(1)
Bond yield (10.298) (3.502) (2.892) 0.000
Domestic lending (3.497) (3.502) (2.893) 0.004
Spot exchange rate (7.468) (3.502) (2.892) 0.000
GDP growth rate (8.513) (3.502) (2.892) 0.000
Defcit to GDP ratio (2.778) (3.502) (2.892) 0.065
Stock market (9.340) (3.502) (2.892) 0.000
Infation (9.340) (3.502) (2.894) 0.027
Residual ?10.298 (3.502) (2.894) 0.000
Notes: *ADF test statistics; **critical value at 1 per cent level of signifcance; ***critical value at 5
per cent level of signifcance
Table II.
Optimal lag selection
Lag LR FPE AIC SC HQ
0 NA 1.350 43.195 43.392 43.274
1 1,727.041 17.438 22.720 24.296* 23.355*
2 94.819 14.782 22.535 25.491 23.726
3 93.051 11.582 22.239 26.574 23.985
4 86.383 9.043 21.888 27.603 24.191
5 79.0197* 7.193* 21.482 28.577 24.340
6 53.641 8.825 21.403* 29.878 24.818
Notes: *Signifcance level (5 per cent); LR ? sequential modifed LR test statistic; FPE ? fnal
prediction error; AIC ? Akaike information criterion; SIC ? Schwarz information criterion; HQ ?
Hannan-Quinn information criterion
335
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
among the variables, based on six lag periods. As co-integration is established, it then
opts for a VECM for the analysis of the six pillars of fnancial stability for the data
set[23].
4.4 The VECM
In a VAR model, the endogenous and explanatory variables interact simultaneously
creating an extended information set. AVECMallows for co-integration. Thereafter,
a basic feature of a VECM is that it includes an error correction term (U
t?1
), which is
a one period lag residual term that provides also reversion to the mean. In this case,
this allows one to explore the association between the six pillars of fnancial stability
and any designated response variable, in so far as the reversion restores the system
to a state of long-run equilibrium. As the full model has seven variables, seven
equations can be specifed, where in the system of equations, any of the seven
variables could be dependent. These can be considered as a nexus of relationships.
However, this study is only interested in some aspects of this nexus. Specifcally it has
identifed six aggregates as pillars of fnancial stability. Thereafter, the objective is to
test the possible associations of these six pillars with economic stability (represented by
real GDP) and then, given the historical/theoretical emphasis placed upon it, price
stability (represented by the CPI). As such, a reduced formof the full model is presented,
which specifes only two equations. In equation (1), GDP appears on the left as the
response variable, and in equation (2), infation appears on the left as the response
variable:
Y
t
(
GDPG
)
? U
t?1
? ßY
t?i
(
GDPG
)
? ßY
t?i
(
SMR
)
? ßX
t?i
(
SER
)
? ßX
t?i
(
INF
)
? ßX
t?i
(
DL
)
? ßX
t?i
(
DGDPR
)
? ßX
t?i
(
BMY
)
? Constant ? et
(1)
Table III.
Johansen co-
integration test
Hypothesized no. of CE(s) Eigen value Trace statistic 0.05 critical value P**
None 0.557 220.580 125.615 0.000
At most 1* 0.424 149.825 95.754 0.000
At most 2* 0.329 101.882 69.819 0.000
At most 3* 0.292 67.118 47.856 0.000
At most 4* 0.219 37.035 29.797 0.006
At most 5* 0.156 15.579 15.495 0.049
At most 6 0.009 0.817 3.841 0.366
Unrestricted co-integration rank test (maximum eigen value)
None 0.557 70.754 46.231 0.000
At most 1* 0.424 47.944 40.078 0.005
At most 2* 0.329 34.764 33.877 0.039
At most 3* 0.292 30.083 27.584 0.023
At most 4* 0.219 21.456 21.132 0.045
At most 5* 0.156 14.762 14.265 0.042
At most 6 0.009 0.817 3.841 0.366
Notes: *Hypothesis of no co-integration was rejected by trace and max eigen value
test; **MacKinnon-Haug-Michelis (1999) P-values
JFEP
7,4
336
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Y
t
(
INF
)
? U
t?1
? ßY
t?i
(
INF
)
? ßY
t?i
(
SMR
)
? ßX
t?i
(
SER
)
? ßX
t?i
(
GDPG
)
? ßX
t?i
(
DL
)
? ßX
t?i
(
DGDPR
)
? ßX
t?i
(
BMY
)
? Constant ? et
(2)
The symbols follow the nomenclature set out in §3; in addition, Y
t
and X
t
refer to the
(n ? 1) vector of time-series endogenous variables, ß
i
refers to the (n ? n) coeffcient
matrixes and et refers to the (n ? 1) white noise or unobservable vector process
(assuming an absence of autocorrelation and the existence of an independent
distribution, i.e. et ?N (0, ?2)).
4.5 Results for VEC equation (1); economic stability
In a VECM, the coeffcients of the error correction term contain information about
whether the past values affect the current values of the variable under study. A
signifcant coeffcient implies that past equilibriumerrors play a role in determining the
current outcomes. The information obtained fromthe error correction model is related to
the speed of adjustment of the system toward long-run equilibrium. The short-run
dynamics are captured through the individual coeffcients of the difference terms.
The results are presented in Table IV: the adjustment coeffcient on the error
correction termfromequation (1) is negative and statistically signifcant at the 5 per cent
level of signifcance. This indicates that when deviating fromthe long-termequilibrium,
the error correction term has an opposite adjustment effect and the degree of deviation
is reduced. In the model, there are six independent variables (representing aspects of
fnancial stability) and a further dependent variable (economic growth). The important
point to highlight here is that the signifcant error term supports the existence of a
long-termrelationship between fnancial stability and economic growth. The R-squared
statistic for the model following equation (1) is 87 per cent. This indicates the strength of
the model in the long run and underpins the signifcance of independent variables and
their lagged effect.
Clearly, then the fndings tend to support more specifc fndings, claims and analyses
that suggest that fnancial stability is a key component in long-term economic
stability[24]; given that price stability cannot guarantee fnancial stability, it then
follows that, a focus on price stability alone will tend to create the potential for issues to
arise in various aspects of fnance that can then pose problems for economic stability
(White, 2006; Borio and White, 2004). These are more than threshold issues in the
Reinhart and Rogoff (2009) sense, as the actual mechanisms contributing to the
manifestation of crisis are likely in some sense to be operative prior to that threshold. For
further analysis, see the results of the IRF in Section 4.52.
4.5.1 Diagnostic tests for equation (1). To check the robustness of this model against
issues of heteroskedasticity, autocorrelation and exogeneity, a series of diagnostic tests
are performed. The results are shown in Table V.
As presented in Table V, the diagnostic test results show that the null of
homoskedasticity (White test) and null of no serial correlation (BG test) could not be
rejected at 5 per cent level of signifcance, which implies that the model and results are
non-spurious. In addition, with the exception of sovereign debt (DGDPR), the variables
showed an exogenous association with economic growth (GDPG) at the 5 per cent level
337
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Table IV.
Vector error
correction model
(VECM): equation (1)
Variables Coeffcient Standard error t-statistic P
U t-1 0.001 0.001 1.085 0.285
U t-1 ?11.378 3.200 ?3.556 0.001
U t-1 ?0.137 0.047 ?2.938 0.006
U t-1 ?0.831 0.354 ?2.348 0.024
U t-1 0.040 0.027 1.514 0.138
U t-1 ?0.308 0.270 ?1.141 0.261
SMR 0.000 0.001 ?0.575 0.568
SMR 0.000 0.001 ?0.257 0.799
SMR 0.000 0.001 0.644 0.523
SMR 0.000 0.000 ?0.812 0.422
SMR 0.000 0.000 0.081 0.936
SMR 0.001 0.000 2.711 0.010
SER 9.130 3.135 2.912 0.006
SER 10.349 2.814 3.678 0.001
SER 8.494 2.386 3.560 0.001
SER 2.137 2.370 0.902 0.373
SER 4.217 1.864 2.262 0.030
SER 2.067 1.659 1.246 0.220
INF 0.107 0.045 2.381 0.022
INF 0.024 0.048 0.500 0.620
INF 0.022 0.038 0.569 0.572
INF 0.031 0.032 0.962 0.342
INF 0.030 0.032 0.924 0.361
INF 0.047 0.038 1.233 0.225
GDPG 0.216 0.344 0.629 0.533
GDPG 0.232 0.305 0.759 0.453
GDPG 0.098 0.230 0.425 0.673
GDPG ?0.157 0.217 ?0.725 0.473
GDPG 0.029 0.191 0.152 0.880
GDPG 0.188 0.159 1.184 0.244
DL 0.448 0.186 2.415 0.021
DL 0.572 0.208 2.753 0.009
DL ?0.304 0.135 ?2.260 0.030
DL ?0.061 0.138 ?0.440 0.663
DL 0.021 0.131 0.158 0.876
DL ?0.088 0.128 ?0.686 0.497
DGDPR ?0.380 0.408 ?0.931 0.358
DGDPR 0.473 0.416 1.137 0.263
DGDPR 0.088 0.391 0.226 0.823
DGDPR ?0.586 0.356 ?1.644 0.108
DGDPR 0.067 0.360 0.186 0.853
DGDPR ?0.017 0.373 ?0.046 0.964
BMY ?2.161 0.874 ?2.471 0.018
BMY ?2.007 0.704 ?2.851 0.007
BMY ?2.107 0.637 ?3.306 0.002
BMY ?1.074 0.621 ?1.729 0.092
BMY ?0.855 0.523 ?1.635 0.110
BMY ?1.040 0.479 ?2.170 0.036
Constant ?2.659 0.876 ?3.036 0.004
Note: Estimation using ordinary least square (OLS) method
JFEP
7,4
338
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
of signifcance. Interestingly, on the whole, all fnancial aggregates (see fnal row) were
signifcantly exogenous to economic growth.
In a VECMwith lags, there are some coeffcients that are statistically insignifcant. A
Wald test was performed to ascertain whether the various explanatory variables and
their coeffcients jointly infuence response variables. The results are shown in Table VI.
As presented in Table VI, all explanatory variables showed an overall signifcant
association with economic growth (GDPG), except sovereign debt (DGDPR), which did
not meet the 5 per cent level of signifcance benchmark.
4.5.2 IRF and data analysis for equation (1). As previously noted, the empirical
results obtained from the VECM and presented in Section 4.5 (Table IV) provide some
insight regarding economic stability but also involve an inherent limitation. To address
this, an IRF analysis is provided. It sets out an IRF for the response of GDP growth
(GDPG) to a one standard deviation shock in each of the six separate fnancial pillars.
The fndings are presented in Figure 1.
It should be noted here that there are no confdence bands around the impulse line
when using a VECMwith error correction terms. This is due to the reason that the error
correction termbrings the systemto the long-termequilibriumand thus no uncertainty
around the mean. As such, there are no logical grounds for confdence interval bands.
However, to overcome this issue, bootstrapping by using the Efron percentile confdence
interval[25] and performing 1000 bootstrap replications (B ?1,000), which is identifed
as a fairly common and established practice in the literature, is adopted.
The key fndings for the UK based on the data set include:
• Aone standard deviation shock to the stock market (SMR) leads to an increase in
economic growth (GDPG) that persisted for more than eight quarters. This
indicates a strong association between the stock market and the real economy in
a non-crisis regime situation. One can then infer a strong impact in terms of a
wealth effect here.
• A one standard deviation shock to the exchange rate (SER) results in a relatively
minor immediate positive impact on economic growth (GDPG), followed by a
subsequent more pronounced and increasing negative impact. Here, one can draw
Table V.
Diagnostic test
(heteroskedasticity,
autocorrelation and
exogeneity)
Heteroskedasticity: White test Test stat P value
Observed R-squared 35.400 P chi-square (48) 0.911
Breusch-Godfrey serial correlation LM test
Observed R-squared 5.867 P chi-square (2) 0.053
Block exogeneity Wald test
D(SMR) 18.567 df-6 0.005*
D(SER) 22.109 df-6 0.001*
D(INF) 12.735 df-6 0.047**
D(DL) 37.723 df-6 0.000*
D(DGDPR) 5.742 df-6 0.453
D(BMY) 13.323 df-6 0.038**
All 82.715 df-6 0.000*
Notes: *Signifcant at 1 per cent level; **signifcant at 5 per cent level
339
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
the inference that exchange rate appreciation may not provide for long-term real
economic growth. This can be based on a variety of actual mechanisms. For
example, where exports of goods are adversely affected (expressed in the balance
of trade) to a degree that is not offset by the balance on invisibles.
• A one standard deviation shock to infation leads to a continuous and increasing
negative impact on real economic growth (GDPG). This seems to support the
generally recognized claim that infation does not contribute to real economic
growth and one can also drawthe inference that price stability, though clearly not
suffcient for economic stability, is an extremely important constituent of a
situation that contributes to such stability.
• A one standard deviation shock (a pricing decrease) to the bond market yield
(BMY) results in a negative impact on real economic growth that begins to
manifest after three quarters, and this is indicative of both the sensitivity of the
economy to yields and a signifcant lag in transitional effects from fnancial
markets to the real sector.
• A one standard deviation shock to domestic lending (DL) provides a positive
impact on real economic growth, and this persists for several quarters. This
confrms the generally recognized positive role of credit creation within a
Table VI.
Wald test: error
correction model
Test statistic Value df P
SMR
F-statistic 3.094 (6, 38) 0.014*
Chi-square 18.566 6 0.005*
SER
F-statistic 3.684 (6, 38) 0.006*
Chi-square 22.109 6 0.001*
INF
F-statistic 2.122 (6, 38) 0.073
Chi-square 12.735 6 0.047*
DGDPG
F-statistic 1.521 (6, 38) 0.197
Chi-square 9.130 6 0.166
DL
F-statistic 6.287 (6, 38) 0.000*
Chi-square 37.723 6 0.000*
DGDPR
F-statistic 0.956 (6, 38) 0.467
Chi-square 5.741 6 0.452
BMY
F-statistic 2.220 (6, 38) 0.062
Chi-square 13.322 6 0.038*
Notes: *Signifcant at 1 per cent level; **signifcant at 5 per cent level
JFEP
7,4
340
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Figure 1.
Impulse response
function: economic
growth and fnancial
stability
341
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
non-crisis regime. This, of course, does not militate against adverse movements in
absolute scales of debt, leverage and terms of credit that can accumulate during
non-crisis periods and which may undermine the very conditions of fnancial and
economic stability.
• A one standard deviation shock to the sovereign debt level in comparison to
national income (DGDPR) provides no immediate signifcant positive impact on
real economic growth but does result in a delayed yet persistent negative impact
after the eighth quarter. This may be related to the way in which public
expenditure is used (as the actual context is productive capital investment vs
more limited forms of welfare payments with different consequences and possible
multipliers).
4.6 Results for VEC equation (2); price stability
The second aspect of economic stability is price stability and its association with
fnancial stability. The results for VEC equation (2) are presented in Table VII.
In Table VII, the adjustment coeffcient on the error correction termfor equation (2) is
positive and not statistically signifcant at the 5 per cent level of signifcance. The term
is also relatively small and so the error correction function is not tending to restore
equilibrium, implying that there will be disruption to equilibrium in the long run.
However, it might expect other external factors, not accounted for in the model, to help
maintain the long-run equilibrium. The R-squared statistic for equation (2) is 83 per cent,
and this indicates the strength of the model in the long run and underpins the
signifcance of independent variables and their lagged effect. For further analysis, see
the results of the IRF in Section 4.62.
4.6.1 Diagnostic tests for equation (2). As with equation (1), to check the robustness
of this model against issues of heteroskedasticity, autocorrelation and exogeneity, a
series of diagnostic tests are performed. The results are shown in Table VIII.
As presented in Table VIII, the diagnostic test results show that the null of
homoskedasticity (White test) was accepted, while the null of no serial correlation (BG
test) was rejected at the 5 per cent level of signifcance. Though the model and results
may be deemed non-spurious, it has signs of serial correlation, which is often the case
with fnancial observations. With the exception of economic growth (GDPG), the
variables did not showsignifcant exogenous association with infation (INF) at the 5 per
cent level of signifcance. Interestingly, on the whole all (see fnal row), fnancial
aggregates were signifcantly exogenous to price stability.
As with equation (1), a Wald test was performed to ascertain whether the various
explanatory variables and their coeffcients jointly infuence response variables. The
results are shown in Table IX.
As presented in Table IX, none of the explanatory variables showed an overall
signifcant association with infation (INF), except economic growth (DGDPG), which
did meet the 5 per cent level of signifcance benchmark. This allows us to infer that
among the six pillars of fnancial stability, economic growth may have a stronger
association with price stability than with the other specifed fnancial sector aggregates.
4.6.2 IRF and data analysis for equation (2). Following the format of Section 4.52, it
sets out an IRF for the response of price stability (INF) to a one standard deviation shock
JFEP
7,4
342
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Table VII.
Vector error
correction model
(VECM): equation (2)
Variables Coeffcient Standard error t-statistic P
U t-1 0.003 0.003 1.095 0.281
U t-1 11.773 12.310 0.956 0.345
U t-1 ?0.293 0.180 ?1.630 0.111
U t-1 ?2.932 1.361 ?2.154 0.038
Ut-1 0.189 0.103 1.841 0.073
U t-1 ?0.306 1.040 ?0.294 0.770
SMR ?0.002 0.003 ?0.571 0.572
SMR ?0.002 0.003 ?0.878 0.386
SMR ?0.001 0.002 ?0.362 0.720
SMR 0.002 0.002 1.087 0.284
SMR 0.000 0.002 0.182 0.857
SMR ?0.001 0.001 ?0.358 0.722
SER ?6.250 12.060 ?0.518 0.607
SER ?4.212 10.823 ?0.389 0.699
SER ?18.580 9.179 ?2.024 0.050
SER ?5.856 9.117 ?0.642 0.525
SER ?2.008 7.170 ?0.280 0.781
SER ?7.867 6.380 ?1.233 0.225
INF 0.090 0.173 0.522 0.605
INF ?0.006 0.185 ?0.030 0.976
INF ?0.362 0.146 ?2.473 0.018
INF 0.244 0.123 1.988 0.054
INF ?0.508 0.124 ?4.115 0.000
INF ?0.436 0.146 ?2.998 0.005
GDPG 2.433 1.322 1.840 0.074
GDPG 0.244 1.174 0.208 0.836
GDPG 0.181 0.886 0.205 0.839
GDPG 0.404 0.835 0.484 0.631
GDPG ?0.117 0.734 ?0.159 0.874
GDPG 0.092 0.613 0.151 0.881
DL ?0.445 0.714 ?0.623 0.537
DL ?0.101 0.799 ?0.127 0.900
DL 0.994 0.518 1.920 0.062
DL 0.104 0.531 0.197 0.845
DL ?0.558 0.504 ?1.107 0.275
DL ?0.379 0.491 ?0.771 0.446
DGDPR 1.519 1.568 0.968 0.339
DGDPR ?1.666 1.600 ?1.041 0.304
DGDPR ?1.420 1.504 ?0.944 0.351
DGDPR ?0.078 1.370 ?0.057 0.955
DGDPR 1.063 1.385 0.767 0.448
DGDPR ?2.079 1.436 ?1.448 0.156
BMY ?3.765 3.364 ?1.119 0.270
BMY ?0.212 2.708 ?0.078 0.938
BMY 1.610 2.451 0.657 0.515
BMY ?0.599 2.390 ?0.251 0.803
BMY ?0.506 2.012 ?0.252 0.803
BMY 0.150 1.843 0.081 0.936
Constant 10.714 3.369 3.180 0.003
Note: Estimation using ordinary least square (OLS) method
343
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Table IX.
Wald test: error
correction model
Test statistic Value df P
SMR
F-statistic 1.064 (6, 38) 0.400
Chi-square 6.388 6 0.3811
SER
F-statistic 1.409 (6, 38) 0.236
Chi-square 8.454 6 0.206
INF
F-statistic 8.216 (6, 38) 0.000*
Chi-square 29.301 6 0.000*
DGDPG
F-statistic 3.127 (6, 38) 0.013*
Chi-square 18.765 6 0.004*
DL
F-statistic 1.413 (6, 38) 0.234
Chi-square 8.483 6 0.204
DGDPR
F-statistic 0.820 (6, 38) 0.561
Chi-square 4.923 6 0.553
BMY
F-statistic 7.122 (6, 38) 0.333
Chi-square 7.122 6 0.309
Notes: *Signifcant at 1 per cent level; **signifcant at 5 per cent level
Table VIII.
Diagnostic test
(heteroskedasticity,
autocorrelation and
exogeneity)
Heteroskedasticity: White test Test stat P value
Observed R-squared 48.763 P chi-square (48) 0.449
Breusch-Godfrey serial correlation LM test
Observed R-squared 13.927 P chi-square (2) 0.001
Block exogeneity Wald test
D(SMR) 6.389 df-6 0.381
D(SER) 8.454 df-6 0.207
D(GDPG) 18.766 df-6 0.005
D(DL) 8.484 df-6 0.205
D(DGDPR) 4.923 df-6 0.554
D(BMY) 7.123 df-6 0.310
All 80.520 df-6 0.000
Notes: *Signifcant at 1 per cent level; **signifcant at 5 per cent level
JFEP
7,4
344
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
in each of the six separate fnancial pillars. The bootstrapping was performed to
construct confdence interval by using the Efron percentile confdence interval method.
One thousand bootstrap replications were carried out (B ? 1,000). The fndings are
presented in Figure 2.
The key fndings for the UK based on the data set include:
• If referring to Figures 2a (the stock market) and 2b (the exchange rate), in both
cases, a one standard deviation shock is associated with a slow and persistent
increase in infation. This fnding accords with the general economic insight
that exchange rate appreciation affects purchasing power and so the basis of
consumption, and that fnancial assets and instruments can provide hedges
against infation and so such assets rise in relation to other sources of price
pressures (and in terms of trade-offs with capital fows to bond markets).
• If referring to Figure 2c (the bond market, BMY), a one standard deviation
shock is also associated with a general and persistent increase in infation.
This fnding is of great interest because it implies many potential points of
linkage between sovereign debt activity and general price levels. Rising
infation may imply a demand among bond investors for higher rates of
nominal return on assets to recover real rates. Conversely, the government
may be involved in tacitly encouraging a degree of infation to reduce its
long-term debt burden based on historic accumulation of treasuries. It may
also be that as the government’s cost of borrowing increases, taxation
(perhaps in regressive stealth forms) also increases, resulting in the
possibility for an additional price pressure.
• If referring to Figure 2d (real economic growth, GDPG), a one standard
deviation shock leads to a decrease in infation. This, of course, is conditional
on appropriate relations to output capacity, based also on investment,
productivity and wage growth.
• Finally, if referring to Figures 2e (domestic lending, DL) and 2f (the sovereign
debt level, DGDPR)[26], a one standard deviation shock leads to divergent
persistent effects on infation between the cases. It appears curious that rising
domestic lending is not clearly associated with rising prices, while sovereign
debt does appear to be. Intuitively, domestic lending is associated with
consumption and also with asset appreciation in given markets, it then seems
worthy of additional research to explore the relations at work here. This,
however, is beyond the scope of this paper.
What should be emphasized here is that the complexity of economic relations
implies that price stability is always under pressure within a non-crisis regime. This
implies that in addition to the point that one cannot assume that price stability will
be suffcient for fnancial stability, one also can neither simply assume the
continuation of price stability nor that fnancial stability will persist in a way that
does not undermine price stability. All these points tend to indicate there is a role for
state oversight and intervention within a non-crisis regime based on the complexity
of possible interactions that may undermine fnancial stability and price stability,
with consequences for their association with economic growth.
345
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Figure 2.
Impulse response
function: price
stability and
fnancial stability
JFEP
7,4
346
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
5. Conclusion: policy implications
This paper has pursued an empirical exploration of six pillars of fnancial stability,
based on a data set for the UK extending from 1985 (Q1) to 2008 (Q2). It has been done
so through the construction of a VECM, including an IRF. The purpose of this paper has
been to provide a different context for considering issues of fnancial stability and
instability, with reference to economic growth and price stability in particular. This
context is more inclusive than the received approach of researchers such as Reinhart and
Rogoff (2009), which creates an inadvertent focus on crisis thresholds. The model has
allowed us to test aspects of fnancial stability against economic and price stability in a
non-crisis situation. The fndings should be considered in the context of a concept of the
economy as fundamentally dynamic and subject to complex cumulative processes.
While it focuses on a non-crisis situation in terms of the data set, it should be aware that
what occurs in such a situation contributes to any subsequent crisis period. As such,
though one can defne and distinguish stability and instability (as it has been done), the
distinction is one that involves some interface ambiguity as well as contingency. This
ambiguity, however, far from providing an argument against implications for policy
tends rather to provide support for cautionary and prudential approaches.
In so far as there is an association between economic growth, fnancial and price
stability and in so far as fnancial instability creates adverse effects, then it would be
argued the fndings provide one justifcation (among many others)[27] for
macro-prudential approaches to the fnance system (White, 2009; Tsouma, 2009;
Mishkin, 2011). It is no longer tenable to argue that particular asset bubbles cannot
be identifed prior to their collapse, nor that it is simply more expedient to allow
them to collapse and deal with the damage thereafter. A modern economy has
multiple points of connection and any given asset appreciation phenomena may be
related to or have consequences for other aspects of an economy (fnance and risk
dispersion can quickly become damaging contagion). This is being increasingly
recognized. For example, it is explicitly stated in the neworganizational architecture
of the Bank of England through the Financial Services Act of 2012 and the creation
of the Financial Policy Committee, whose new remit is expressly stated as
macro-prudential (Osborne, 2013). The specifc form that this remit will then
manifest in is yet to be set; macro-prudential policy is an evolving issue. However, it
clearly makes sense to provide for oversight and regulation that begins from the
potential problems that may arise in non-crisis periods and to consider these based
on particular sets of associations that can then be explored in more detail. This
paper has made a start here in identifying some of these.
Notes
1. Though in the context of a longer tradition of events-focused modeling, see Kaminsky and
Reinhart (1999), Honohan (1997) and contrast with Hardy and Pazarbasioglu (1998).
2. Though the purpose here is to create a different approach to thresholds.
3. Note: A period of relative stability is not also to be confated with an absence of cumulative
problems. The great moderation, for example, was ostensibly a period of price and, to some
degree, fnancial stability, but is also widely criticized because it was also a period of
cumulative changes, which resulted in fnancial crisis (see Borio, 2011 and Section 2
hereafter).
347
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
4. Of course “without giving way to cumulative processes” does not imply that there are no
qualitative and quantitative changes occurring that may become cumulative and
problematic in the future – stability is a conditional feature and not a fxed constituent;
otherwise, the defnition would be perverse, as periods of stability could only become
periods of serious instability due to exogenous x factors rather than endogenous
processes.
5. Mishkin (1991) provides a similar generic defnition.
6. Even Fred Mishkin (2011, pp. 30-31), who prior to the Global Financial Crisis was
considered the main orthodox proponent of narrowly focused central bank policy, now
states: “The price and output stability do not ensure fnancial stability. Policy focus
solely on these (output, infation) objectives may not be enough to produce good
economic outcomes”. However, there is some dissent regarding this claim (Nakov and
Thomas, 2011).
7. Bordo et al. (2003) construct an Annual Index of Financial Conditions based on the
categories: severe fnancial distress, moderate distress, normal, fnancial expansion and
fnancial euphoria. The index was based on bankruptcies, corporate insolvencies and
asset prices. They fnd that the impact has generally decreased in the post Second World
War period, perhaps because of (until the 1990s) more effective regulation. Most
important point here is that they were looking in the one direction, which was impact of
price level on fnancial instability not the other way round.
8. According to Martin and Milas (2012), monetary policy has signifcant impacts on infation
and output but these do not persist during periods of manifest crisis, as the focus of policy
responses shifts toward the causes of fnancial stress.
9. The study focuses on fnancial stress using the International Monetary Fund’s (IMF)
Financial Stress Index (FSI). The FSI has three subcomponents: the banking sector (the slope
of the yield curve, TED spread and the beta of banking-sector stocks), securities markets
(corporate bond spreads, stock-market returns and time-varying volatility of stock returns)
and exchange rates (time varying volatility of NEER changes).
10. One might also note that central bank policy is evolving. The Bank of England (2008) has
published a biannual systemic risk survey, which canvases the opinion of risk directors
among a sample of hedge funds, banks, building societies, asset managers and insurers to
construct a snapshot of the general foci of contemporary concern: geopolitical risk, sovereign
risk, operational risk, etc. (see Appendix).
11. Data set available as xls format from: www.bankofengland.co.uk/statistics/Pages/
yieldcurve/archive.aspx
12. Data set available as xls format from:http://data.worldbank.org/indicator/FS.AST.PRVT.
GD.ZS
13. Specifcally: “fnancial resources provided to the private sector by fnancial corporations,
such as through loans, purchases of non-equity securities, and trade credits and other
accounts receivable, that establish a claimfor repayment. The fnancial corporations include
monetary authorities and deposit money banks, as well as other fnancial corporations
where data are available (including corporations that do not accept transferable deposits but
do incur such liabilities as time and savings deposits). Examples of other fnancial
corporations are fnance and leasing companies, money lenders, insurance corporations,
JFEP
7,4
348
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
pension funds, and foreign exchange companies.”http://data.worldbank.org/indicator/FS.A
ST.PRVT.GD.ZS
14. See also Albero (2011).
15. Adjusted for dividends and splits. The data set is publically available via several platforms.
See www.ftse.com/analytics/factsheets/Home/HistoricIndexValues orhttps://uk.fnance.
yahoo.com/q/hp?s?%5EFTSE
16. The data set is available at: www.ons.gov.uk/ons/datasets-and-tables/data-selector.html?
cdid?CDKO&dataset?mm23&table-id?3.5
17. For the calculations and periodic changes to the weighting system, see www.ons.gov.uk/
ons/guide-method/user-guidance/prices/cpi-and-rpi/cpi-and-rpi–updating-weights/index.
html. See the Royal Statistical Society (RSS) working party analysis of the key differentials
between RPI and CPI, Leyland (2011).
18. The data set is available at: www.bankofengland.co.uk/boeapps/iadb/index.asp?Travel?N
IxIRx&levels?1&XNotes?Y&C?DLM&G0Xtop.x?17&G0Xtop.y?6&FNotes2?Y&XN
otes2?Y&Nodes?X3790X3791X3873X33940X3836&SectionRequired?I&HideNums?-1
&ExtraInfo?true#BM
19. Note: it is not an inconsistency to use forward data for bond yields and spot data for exchange
rates, as this study is interested in different relations in terms of each. Expectations regarding
yield are a signifcant factor for decision making in terms of bond purchases, while this is not a
core issue for currencies, hence spot data are appropriate.
20. The data set is available at: www.ons.gov.uk/ons/datasets-and-tables/data-selector.html?
cdid?J4DE&dataset?pusf&table-id?PSF9
21. The data set is available at: www.ons.gov.uk/ons/datasets-and-tables/data-selector.html?
cdid?KGX4&dataset?pn2&table-id?C1
22. So one arisingissue is that multiple lags inthe independent or explanatoryvariables prohibit
a complete picture of the associations among the said variables.
23. Note: Elliott (1998) highlights the robustness of co-integration methods when series are very
close to unit root and fnds that hypothesis tests are unaffected by the presence of near unit
root variables not included in the restrictions.
24. See also Beck et al. (2008), Kunt et al. (2004) and Guiso et al. (2004); each suggests that
through its effect on the allocation of fnancial resources, the quality of bank regulation and
supervision may have dramatic effects on economic growth.
25. It used JMulti-4 software package for bootstrapping; for details on the bootstrapping method
and its rationale, please see Efron and Tibshirani (1993).
26. Recalling, this variable is selected as a defcit to income ratio, rather than selecting a debt to
income ratio, because it provides an ongoing snapshot of the prevailing shortfall or surplus in
government fnances, rather than a cumulative expression of past debt. As such, it provides an
indication of the potential for proximate fnancial distress or its absence in recent events.
27. For example, Caruana (2010) claims that the global fnancial crisis has provided a deeper
insight into the role central banks can and should play in encouraging fnancial system
stability; Giannone et al. (2011) demonstrate that the quality of bank regulation helps to
account for the scale of damages triggered by a fnancial crisis.
349
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
References
Airaudo, M., Nistico, S. and Zanna, L.F. (2008), “Learning monetary policy and asset prices”,
Working Paper No. 48, LLEE.
Albero, C.F. (2011), Financial Frictions, the Financial Immoderation, and The Great Moderation,
Department of Economics, Rutgers University.
Arouri, M., Jawadi, F. and Nguyen, D.K. (2013), “What can we tell about monetary policy
synchronization and interdependence over the 2007-2009 global fnancial crisis?”, Journal
of Macroeconomics, Vol. 36 No. C, pp. 175-187.
Bagehot, W. (1873), Lombard Street: A Description of the Money Market, John Murray, London,
Reprinted with Introduction by Hartley Withers 1924, William Clowes and Sons, London.
Bank of England (2008), Financial Stability Report, No. 23, Working Paper 2001-018A, available at:
www.bankofengland.co.uk/publications/Pages/fsr/default.aspx (accessed 10 December
2012).
Basu, S. and Michailidis, G. (2013), Estimation in High-dimensional Vector Autoregressive Models,
Mathematics, Statistics Theory, Cornell University Library, Ithaca.
Baxa, J., Horvath, R. and Vasicek, B. (2013), “Time-varying monetary-policy rules and fnancial
stress: does fnancial instability matter for monetary policy?”, Journal of Financial Stability,
Vol. 9 No. 1, pp. 117-138.
Beck, T., Demirguc-Kunt, A., Laeven, L. and Levine, R. (2008), “Finance, frm size and growth”,
Journal of Money, Credit and Banking, Vol. 40 No. 7, pp. 1379-1405.
Blanchard, O., Dell’Ariccia, G. and Mauro, P. (2010), “Rethinking macroeconomic policy”, IMF
Staff Position Note, SPN/10/03, Working Paper 2001-018A, available at: www.imf.org/
external/pubs/ft/spn/2010/spn1003.pdf (accessed 30 January 2013).
Bordo, M.D., Dueker, M.J. and Wheelock, D.C. (2003), “Aggregate price shocks and fnancial
stability: the United Kingdom 1796-1999”, Working Paper 2001-018A, available at: http://
research.stlouisfed.org/wp/2001/2001-018.pdf (accessed 7 December 2013).
Borio, C. (2011), “Rediscovering the macroeconomic roots of fnancial stability policy: journey,
challenges and a way forward”, Working Paper No. 354, BIS.
Borio, C. and White, W. (2004), “Whither monetary and fnancial stability? The implications of
evolving policy regimes?”, Working Paper No. 147, BIS.
Brooks, C. (2008), Introductory Econometrics for Finance, 2nd ed. Cambridge University Press,
Cambridge.
Campbell, J.Y. (1995), “Some lessons fromthe yield curve”, Journal of Economic Perspectives, Vol. 9
No. 3, pp. 129-152.
Canova, F. (2007), Methods for Applied Macroeconomic Research, Princeton University Press, NJ.
Caruana, J. (2010), Macro Prudential Policy: Working Toward a New Consensus, Bank for
International Settlements, Basel, Switzerland.
Efron, B. and Tibshirani, R. (1993), An Introduction to the Bootstrap, Chapman &Hall/CRC Press,
Boca Raton, FL.
Elliott, G. (1998), “On the robustness of cointegration methods when regressors almost have unit
roots”, Econometrica, Vol. 66 No. 1, pp. 149-158.
Engle, R.F. and Granger, C.W.J. (1987), “Co-integration and error correction: representation,
estimation, and testing”, Econometrica, Vol. 55 No. 2, pp. 251-276.
Filardo, A. (2008), “Household debt, monetary policy and fnancial stability: still searching for a
unifying model”, Working Papers No. 46, BIS.
JFEP
7,4
350
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Foot, M. (2003), “What is fnancial stability and how do we get it?”, ACI (UK) The Roy Bridge
Memorial Lecture, available at: www.fsa.gov.uk/library/communication/speeches/2003/sp
122.shtml (accessed 7 March 2012).
Friedman, M. (1998), “A comment on CSWEP”, Journal of Economic Perspectives, American
Economic Association, Vol. 12 No. 4, pp. 197-199.
Funke, M., Paetz, M. and Pytlarczyk, E. (2011), “Stock market wealth effects in an estimated DSGE
model for Hong Kong”, Economic Modelling, Vol. 28, pp. 316-334.
Gertler, M.L. (1988), “Financial structure and aggregate economic activity: an overview”, Working
Papers No. 2559, National Bureau of Economic Research (NBER).
Giannone, D., Lenza, M. and Reichlin, L. (2011), “Market freedom and the global recession”, IMF
Economic Review, Vol. 59 No. 1, pp. 111-135.
Goodhart, C.A.E. (2004), “Some new directions for fnancial stability?”, Financial Markets Group,
London School of Economics, London.
Greene, W.H. (2012), Econometric Analysis, 7th ed., Pearson, London.
Gros, D. (2010), Monetary Policy, Fiscal Policy and Financial Stability: AFat Hypothesis on Mutual
Linkages, Policy Department Economic and Scientifc Policies, European Parliament,
B-1047, Brussels.
Guiso, L., Jappelli, T., Padula, M. and Pagano, M. (2004), “Financial market integration and
economic growth in the EU”, Economic Policy, Vol. 19 No. 40, pp. 523-577.
Gulley, O.D. and Sultan, J. (2003), “The link between monetary policy and stock and bond markets:
evidence from the federal funds futures contract”, Applied Financial Economics, Vol. 13
No. 3, pp. 199-209.
Hagen, von J. and Ho, T.K. (2007), “Money market pressure and the determinants of banking
crises”, Journal of Money, Credit and Banking, Vol. 39 No. 5, pp. 1037-1066.
Hardy, D. and Pazarbasioglu, C. (1998), “Leading indicators of banking crises: was Asia
different?”, Working Paper No. 98/91, IMF.
Hernden, T., Ash, M. and Pollin, R. (2014), “Does high public debt consistently stife economic
growth? Acritique of Reinhart and Rogoff”, Cambridge Journal of Economics, Vol. 38 No. 2,
pp. 257-279.
Ho, T.K. (2004), “How useful are regime-switching models in banking crises identifcation?”,
Working Paper No. 764, Far Eastern Meetings, Econometric Society.
Honohan, P. (1997), “Banking system failures in developing and transition countries: diagnosis
and prediction”, Working Paper No. 39, BIS.
Kaminsky, G. and Reinhart, C. (1999), “The twin crises: the causes of banking and
balance-of-payments problems”, American Economic Review, Vol. 89 No. 3, pp. 473-500.
Khorasgani, H.S. (2010), “Financial instability and optimal monetary policy rule”, Working Paper
No. 42, FIW.
Kindleberger, C.P. (1978), Manias, Panics, and Crashes: A History of Financial Crises, Macmillan,
London.
Kunt, D.A., Love, I. and Maksimovic, V. (2004), Business Environment and the Incorporation
Decision, Policy Research Working Paper Series 3317, The World Bank.
Leyland, J. (2011), “RPI versus CPI – the defnitive account”, available at: www.statslife.org.uk/
economics-and-business/38-rpi-versus-cpi-the-defnitive-account (accessed 7 May 2014).
Liew, V.K.S. (2004), “Which lag length selection criteria should we employ?”, Economics Bulletin,
Vol. 3 No. 33, pp. 1-9.
351
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
MacKinnon, J.G., Haug, A.A. and Michelis, L. (1999), “Numerical distribution functions of
likelihood ratio tests for cointegration”, Journal of Applied Econometrics, Vol. 149 No. 5,
pp. 563-577.
Malikane, C. and Semmler, W. (2008), “Asset prices, output and monetary policy in a small open
economy”, Metroeconomica, Vol. 59 No. 4, pp. 666-686.
Martin, C. and Milas, C. (2012), “Financial crises and monetary policy: evidence from the UK”,
Journal of Financial Stability, Vol. 9 No. 4, pp. 654-661.
Minsky, H.P. (1974), “The modeling of fnancial instability: an introduction”, Modeling and
Simulation, Conference Proceedings of the Fifth Annual Pittsburgh Conference, Pittsburgh.
Minsky, H.P. (1982), “The federal reserve between a rock and a hard place”, in Can It Happen
Again? Essay on Instability and Finance, M.E. Sharpe, Armonk, NY, pp. 192-202.
Mishkin, F.S. (1991), Anatomy of a Financial Crisis, NBERWorking Papers 3934, National Bureau
of Economic Research, Inc.
Mishkin, F.S. (2011), “Over the cliff: from the subprime to the global fnancial crisis”, Journal of
Economic Perspectives, Vol. 25 No. 1, pp. 9-70.
Muscatelli, A., Tirelli, P. and Trecroci, C. (2004), “Fiscal and monetary policy interactions:
empirical evidence and optimal policy using a structural New-Keynesian model”, Journal of
Macroeconomics, Vol. 26 No. 2, pp. 257-280.
Nakov, A. and Thomas, C. (2011), “Optimal monetary policy with state-dependent pricing”,
Working Paper No. 1130, Bank of Spain, Spain.
Nasir, M.A. and Soliman, A.M. (2014), “Aspect of policy combination & effects on fnancial
markets”, Economic Issues, Vol. 19, pp. 92-118.
Osborne, G. (2013), Remit and Recommendations for the Financial Policy Committee, HM
Treasury, April.
Ostry, J.D., Ghosh, A.R., Chamon, M. and Qureshi, M.S. (2012), “Tools for managing
fnancial-stability risks from capital infows”, Journal of International Economics, Vol. 88
No. 2, pp. 407-421.
Pecican, E.S. (2010), “Forecasting based on open VAR model”, Romanian Journal of Economic
Forecasting, Vol. 13 No. 1, pp. 59-69.
Reinhart, C. and Rogoff, K. (2009), “This time it’s different: eight centuries of fnancial
folly-preface”, MPRA Paper 17451, University Library of Munich, Germany.
Schioppa, T.P. (2002), “Central banks and fnancial stability: exploring a land in between”, paper
presented at the Second ECB Central Banking Conference, Frankfurt am Main, 24-25
October.
Tsouma, K. (2009), “Stock returns and economic activity in mature and emerging markets”, The
Quarterly Review of Economics and Finance, Vol. 49 No. 2, pp. 668-685.
White, W.R. (2006), “Is price stability enough?”, Working Paper No. 205, BIS.
White, W.R. (2009), “Should monetary policy ‘Lean or Clean’? Federal Reserve Bank of Dallas
Globalization and Monetary Policy Institute”, Working Paper No. 34, BIS.
Further reading
Howard, D. and Green, D. (2010), Banking on the Future: The Fall and Rise of Central Banking,
Princeton University Press, NJ.
JFEP
7,4
352
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Appendix
Corresponding author
Muhammad Ali Nasir can be contacted at: [email protected]
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]
Figure A1.
Biannual systemic
risk survey
353
Financial and
economic
stability
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
:
5
3

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)

doc_962535558.pdf