Description
The purpose of this paper is to investigate empirically contagion channels of the 2007 US
subprime financial crisis by employing a multivariate GARCH model for four major, international
equity markets, namely the USA, EMU, China and Japan.
Journal of Financial Economic Policy
Contagion channels of the USA subprime financial crisis: Evidence from USA, EMU,
China and Japan equity markets
Dimitrios Dimitriou Theodore Simos
Article information:
To cite this document:
Dimitrios Dimitriou Theodore Simos, (2013),"Contagion channels of the USA subprime financial crisis",
J ournal of Financial Economic Policy, Vol. 5 Iss 1 pp. 61 - 71
Permanent link to this document:http://dx.doi.org/10.1108/17576381311317781
Downloaded on: 24 January 2016, At: 21:46 (PT)
References: this document contains references to 21 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 288 times since 2013*
Users who downloaded this article also downloaded:
Salman Khan, Pierre Batteau, (2012),"Government intervention in Russian bourse: a case
of financial contagion", J ournal of Financial Economic Policy, Vol. 4 Iss 4 pp. 320-339 http://
dx.doi.org/10.1108/17576381211279299
Linyue Li, Nan Zhang, Thomas D. Willett, (2012),"Measuring macroeconomic and financial market
interdependence: a critical survey", J ournal of Financial Economic Policy, Vol. 4 Iss 2 pp. 128-145 http://
dx.doi.org/10.1108/17576381211228989
Simplice A. Asongu, (2012),"The 2011 J apanese earthquake, tsunami and nuclear crisis: Evidence of
contagion from international financial markets", J ournal of Financial Economic Policy, Vol. 4 Iss 4 pp.
340-353http://dx.doi.org/10.1108/17576381211279307
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Contagion channels of the USA
subprime ?nancial crisis
Evidence from USA, EMU, China and
Japan equity markets
Dimitrios Dimitriou and Theodore Simos
Department of Economics, University of Ioannina,
Ioannina, Greece
Abstract
Purpose – The purpose of this paper is to investigate empirically contagion channels of the 2007 US
subprime ?nancial crisis by employing a multivariate GARCH model for four major, international
equity markets, namely the USA, EMU, China and Japan.
Design/methodology/approach – In this study, contagion channels of the 2007 US subprime
?nancial crisis are investigated empirically by employing a multivariate GARCH model for four major,
international equity markets, namely the USA, EMU, China and Japan.
Findings – There is empirical evidence of contagion in all markets with the US market through
various channels, which have not been discussed in other related studies. Speci?cally, the empirical
results suggest that Japanese and EMU markets have been directly affected from the crisis. However,
while China’s equity market has been mainly unaffected by the US subprime crisis, has been affected
indirectly through Japan. Moreover, the Japanese equity market exhibits positive and signi?cant
spillovers effects with China and EMU, revealing an indirect volatility transmission channel of US
subprime crisis.
Research limitations/implications – Further research could consider the asymmetric effects
on conditional covariance through, for example, asymmetric generalized dynamic conditional
correlation models. All under examination markets show evidence of contagion through different
channels.
Practical implications – Despite the ?nancial advices for diversi?cation, since the increasing
globalization and stock market interdependence throughout the last 15 years, through the US subprime
crisis equity investors hadfewer opportunities for diversi?cation. Frompolicy makers’ perspective, they
should carefully examine and uncover possible decoupling strategies to insulate these economies from
contagion in future crises.
Social implications – This study provides useful information to international organizations, such
as World Bank and World Trade Organization (WTO) in order to protect markets from contagion
during future crises.
Originality/value – A novel ?nding of this paper is the indirect channel of contagion (i.e. Japanese
market) for Chinese market. This indirect channel may help explain why China’s equity market
performed badly in 2008 after the subprime crisis in the USA emerged.
Keywords United States of America, Japan, China, European Monetary Union, National economy,
Stock markets, Contagion channels, USA subprime crisis; M-GARCH models
Paper type Research paper
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JEL classi?cations – F3, F36, G01
We are grateful to two anonymous referees of this Journal for providing valuable comments
which substantially improved the quality of this article. Any remaining errors are ours.
Journal of Financial Economic Policy
Vol. 5 No. 1, 2013
pp. 61-71
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381311317781
Contagion
channels
61
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
1. Introduction
The 2007 US subprime ?nancial crisis, due to its severity and consequences to
international markets, is one of the most unanticipated economic events after the Great
Depression of 1930s and attracted great attention from academics, investors and policy
makers.
In this paper we empirically investigate contagion among the US market, the source
of the crisis, two major developed markets Japan and EMU and the most important
emerging market, China. The selection of these markets is based on market
capitalization and GDP in PPP terms. According to CIA World Factbook, at 1 January
2011 these markets are in the ?rst four positions according to the above criteria. Our
sample, weighted MSCI daily data, covers the period from April 1996 to April 2011,
including the 2007 US subprime crisis.
According to Forbes and Rigobon (2002) the contagion hypothesis is de?ned as
“a signi?cant increase in cross-market linkages after a shock to one country (or group
of countries)”, otherwise is considered to be “no contagion, only interdependence”.
Contagion is present during periods of serious worldwide turmoil. Therefore, the
measure of contagion and its transmission channels have been empirical issues under
intense investigation. Prior literature reported several crises transmission mechanisms.
King and Wadhwani (1990) and Kaminsky and Schukler (1999) suggested an analysis
based on revision of expectations[1] and herding[2] behaviour, respectively. In cases
of insulation during period of crises many researchers accept decoupling – recoupling
hypothesis, especially between developed and emerging markets (Dooley and
Hutchison, 2009).
The results from the literature investigating the contagion of US subprime crisis are
mixed. Dooley and Hutchison (2009) provide evidence of contagion on 14 emerging
markets. Kim and Kim (2011) ?nd evidence of ?nancial contagion, around the collapse
of Lehman Brothers in September 2008, on ?ve emerging Asian markets. Furthermore,
Syllignakis and Kouretas (2011) capture contagion effects among USA and German
stock market and seven emerging Central and Eastern Europe markets. On the other
hand, Rose and Spiegel (2009) ?nd little support of contagion using a wide number of
possible causes in a ?exible statistical framework for 107 countries.
Early research on the contagion effects used VAR and simple correlation analysis.
Our study applies a VECH representation of the multivariate GARCH (1,1) model to
estimate the spillover effects. Then we test their signi?cance during pre- and post-crisis
period. According to Ding and Engle (2001), a VECH representation is among the most
parsimonious versions of multivariate GARCH speci?cations[3].
This paper contributes the existent literature on several aspects. We empirically
investigate developed and emerging markets; by examining simultaneously the USA,
EMU, Japan and China stock market spillovers effects. In this way indirect channels of
contagion are revealed (i.e. one market is not affected directly by USA but indirectly
through another market).
Three interesting aspects emerged from our empirical analysis. First, we found
indirect channels of contagion for Chinese market via the Japanese market. As a result
China is indirectly affected by the US subprime crisis. Second, Japan and EMU markets
have been affected directly from the US crisis. Third, EMU directly affected by Japan.
The rest of the paper is structured as follows. In Section 2 we describe the
data employed and their descriptive statistics. The econometric techniques to estimate
JFEP
5,1
62
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the volatility spillovers are stated in Section 3. The empirical results and robustness
tests are discussed in Section 4. Last section concludes.
2. Data description and summary statistics
We employ a daily data set, comprising 3,912 observations for each market, and extend
over 15 years, spanning from April 1996 to April 2011. The data sample covers the US
?nancial crisis initiated in August 2007[4]. However, several other, more localized, crises
occurred during the sample period (e.g. Asian crisis in 1997, Russia’s rouble devaluation
in 1998, and the Brazilian real devaluation in 1999), although we focus mainly on the
recent US subprime due to the severity with which it gripped the markets and
economies globally. The returns are US-dollar denominated[5], computed using
weighted MSCI stock indices fromUSA, EMU, Japan and China equity markets. All data
are extracted from Datastream. The daily stock log-returns are evaluated using r
t
¼ ln
( p
t
/p
t21
) where p
t
is the market total index expressed in US dollars at time t.
Descriptive statistics for the pre-crisis and after-crisis periods are reported in Table I.
The return series appear to be non-normal due to excess kurtosis and fat tails,
supporting the chosen speci?cation for the GARCH innovations. Speci?cally, returns
from the Chinese market have a positive skewness, while returns from USA and EMU
are negatively skewed. Japan’s returns have positive skewness in pre-crisis period and
negative thereafter. The kurtosis, in all time series, exceeds three, indicating a
leptokurtic distribution. Excess kurtosis in equity returns has been well documented by
a number of other studies including Bekaert and Harvey (1997). Test results of the null
hypothesis of normality, using Jarque-Bera statistic, are reported in Table I. With all
USA EMU Japan China
Pre-crisis
Mean 8.72 £ 10
205
0.000167 1.07 £ 10
205
2.63 £ 10
205
Median 0.000410 0.000350 20.000161 20.000205
Maximum 0.053479 0.054348 0.109175 0.098866
Minimum 20.075337 20.057864 20.070541 20.130537
SD 0.010919 0.010002 0.013212 0.018652
Skewness 20.227170 20.222594 0.203966 0.044444
Kurtosis 6.6891 5.132132 6.225434 7.278809
Jarque-Bera 1,698.58
* * *
5,831.38
* * *
1,299.20
* * *
2,251.35
* * *
p-value (0.0000) (0.0000) (0.0000) (0.0000)
During crisis
Mean 26.20 £ 10
205
20.000319 20.000342 0.000181
Median 0.000514 0.000208 0.000000 0.000155
Maximum 0.109019 0.101241 0.104183 0.157124
Minimum 20.094087 20.098206 20.087861 20.135215
SD 0.017381 0.018807 0.017036 0.026602
Skewness 20.222718 20.035154 20.197545 0.234680
Kurtosis 9.54959 7.13294 7.64432 8.62253
Jarque-Bera 1,718.43
* * *
6,813.10
* * *
8,663.18
* * *
1,269.35
* * *
p-value (0.0000) (0.0000) (0.0000) (0.0000)
Notes: Statistically signi?cant at:
*
10,
* *
5 and
* * *
1 per cent levels; the Jarque-Bera LM statistic is
distributed asymptotically as x
2
(2) under the null hypothesis of normality
Table I.
Summary statistics
of daily returns of
international markets
Contagion
channels
63
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
p-values approximately zero to four decimal places, we reject the null hypothesis for all
markets. A visual perspective of the return’s volatility and the impact of 2007 crisis can
be obtained from graphs of stock indices levels, daily returns and conditional variances
for each market in Figure 1. The increase of volatility levels is noticeable in all markets
after the 2007 crisis, revealing clustering effects. Prior less sever crises partially explain
the increased volatility of stock returns during the period 1997-2000.
3. Econometric estimation of the spillover effects
We ?lter out the linear structure of the returns series to decouple them from the
conditional variance. To this end, we employ the VAR model:
r
t
¼ g þ
X
n
s¼1
a
s
r
t2s
þ1
t
; ð1Þ
where r
t
is the 4 £ 1 column vector of equity markets returns, g and a
s
are,
respectively, a 4 £ 1 vector and 4 £ 4 matrices of parameters and 1
t
are 4 £ 1 vectors of
innovations. The lag length, n ¼ 4, is chosen by information criteria[6]. A frequently
employed speci?cation of the conditional variance is Bollerslev et al. (1988)
representation of the multivariate GARCH model. In VECH form, the conditional
covariance matrix is given by:
vechðH
t
Þ ¼ c þ
X
q
j¼1
A
j
vechðu
t2i
u
0
t2i
Þ þ
X
p
j¼1
B
j
vechðH
t2j
Þ; ð2Þ
where vech( · ) denotes the column stacking operator of the lower portion of a symmetric
matrix. So, c is a N(N þ 1)/2 £ 1 vector and matrices A
j
and B
j
are of dimension
N(N þ 1)/2 £ N(N þ 1)/2.
The VECH representation of MGARCH is among the most parsimonious
multivariate GARCH speci?cations. Thus, we employ the diagonal VECH version of
multivariate GARCH (1,1) model[7] (Ding and Engle, 2001):
H
t
¼
CC
0
þ
AA
0
^u
t21
u
0
t21
þ
BB
0
^H
t21
; ð3Þ
Figure 1.
Markets returns,
conditional variance
and levels ?gures
JFEP
5,1
64
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
where u
i;t
jV
t21
, student 2ð0; H
t
; vÞ; V
t21
is the information set at time t 2 1, ^ is
the Hadamard (element by element) product. Since the null hypothesis of normality is
rejected, we assume that the conditional probability density function of the errors
follow the Student’s t-distribution. This is an empirically assumption that takes into
account the fat tail behaviour of the daily data set and, under correct speci?cation, may
improve the ef?ciency of the estimates. Since
CC
0
;
AA
0
and
BB
0
are all positive
semi-de?nite, H
t
will be positive de?nite for all t as far as the initial covariance matrix
H
0
is positive de?nite.
Let v
ij
¼ ð
C; C
0
Þ
ij
, a
ij
¼ ð
AA
0
Þ
ij
, b
ij
¼ ð
BB
0
Þ
ij
then:
h
ijt
¼ v
ij
þa
ij
u
it21
u
jt21
þb
ij
h
ijt21
; i; j ¼ 1; . . . ; N: ð4Þ
The elements of matrix A matrix measure the intensity of spillover effects among
markets. According to Chiang et al. (2007) when the cross-spillover coef?cients of
matrix A are positive, statistical signi?cant and rise after a crisis, we obtain evidence
of contagion. The elements of matrix B measure the persistence of conditional variance.
We estimate the rank one speci?cation of multivariate GARCH (1,1) model which is of
the form:
H
t
¼
CC
0
þ aa
0
^u
t21
u
0
t21
þ bb
0
^H
t21
; ð5Þ
where a and b are N £ 1 vectors. Here we impose the rank of parameter matrices A and
B to be one. The model is estimated using the full information maximum likelihood
(FIML) method with Student t-distributed errors. The FIML estimates are obtained by
maximizing the log-likelihood
P
T
t¼1
l
t
; where:
l
t
¼ log
Gððn þ mÞ=2Þn
m=2
ðnpÞ
m=2
Gðn=2Þðn 22Þ
m=2
2
1
2
logðjH
t
jÞ 2
1
2
ðn þ mÞlog 1 þ
1
0
t
H
21
t
1
t
n 22
" #
; ð6Þ
m is the number of equations, 1
t
is the m vector of residuals and v is the degree of
freedom.
4. Empirical results
At 9 August 2007 the BNP Paribas, a global investment bank, suspended its funds
after deteriorations in subprime mortgage loans. Several researchers used this date as
the starting point namely: Baur (2012), Brunnermeir (2019) and Cecchetti (2009) among
others. Based on this starting date we divide the sample into two sub-periods (pre- and
post-crisis). Estimates of linkages from equation (5), for each sub-period, are reported in
Tables II and III. In line with several other studies (Worthington and Higgs, 2004;
Saleem, 2008) cross-spillover estimates are positive and statistically signi?cant, for
both sample periods, indicating the presence of strong linkages. In Figure 2 we graph
the magnitude of cross-spillover effects for the two sub-periods. In the following
subsection we analyze these results extensively.
4.1 Interpretation of results and robustness tests
The estimated results of cross-spillover effects (Figure 2) indicate a signi?cant
increase, after crisis, for all pairs except of USA-China and EMU-China. These results
indicate direct and indirect channels of contagion for the markets under examination.
Contagion
channels
65
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
U
S
A
(
i
¼
1
)
E
M
U
(
i
¼
2
)
J
a
p
a
n
(
i
¼
3
)
C
h
i
n
a
(
i
¼
4
)
M
a
r
k
e
t
i
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
(
i
,
1
)
1
.
0
1
£
1
0
2
0
9
0
.
7
7
8
4
C
(
i
,
2
)
4
.
1
6
£
1
0
2
0
9
0
.
6
1
8
0
1
.
7
2
£
1
0
2
0
8
0
.
2
5
8
0
C
(
i
,
3
)
8
.
9
6
£
1
0
2
0
9
0
.
5
8
2
8
3
.
7
0
£
1
0
2
0
8
*
0
.
0
8
1
1
7
.
9
6
£
1
0
2
0
8
*
0
.
0
6
3
7
C
(
i
,
4
)
6
.
5
7
£
1
0
2
0
8
0
.
5
7
4
7
2
.
7
£
1
0
2
0
7
*
*
0
.
0
3
0
2
5
£
1
0
2
0
7
*
*
*
0
.
0
0
0
8
4
£
1
0
2
0
6
*
*
*
0
.
0
0
0
0
A
(
i
,
l
)
0
.
0
2
6
3
*
*
*
0
.
0
0
0
0
A
(
i
,
2
)
0
.
0
2
8
1
*
*
*
0
.
0
0
0
0
0
.
0
3
0
0
*
*
*
0
.
0
0
0
0
A
(
i
,
3
)
0
.
0
2
5
2
*
*
*
0
.
0
0
0
0
0
.
0
2
6
9
*
*
*
0
.
0
0
0
0
0
.
0
2
4
1
*
*
*
0
.
0
0
0
0
A
(
i
,
4
)
0
.
0
4
2
7
*
*
*
0
.
0
0
0
0
0
.
0
4
5
6
*
*
*
0
.
0
0
0
0
0
.
0
4
0
8
*
*
*
0
.
0
0
0
0
0
.
0
6
9
1
*
*
*
0
.
0
0
0
0
B
(
i
,
1
)
0
.
9
7
6
0
*
*
*
0
.
0
0
0
0
B
(
i
,
2
)
0
.
9
7
6
8
*
*
*
0
.
0
0
0
0
0
.
9
7
3
2
*
*
*
0
.
0
0
0
0
B
(
i
,
3
)
0
.
9
7
6
8
*
*
*
0
.
0
0
0
0
0
.
9
7
5
5
*
*
*
0
.
0
0
0
0
0
.
9
7
7
7
*
*
*
0
.
0
0
0
0
B
(
i
,
4
)
0
.
9
4
6
8
*
*
*
0
.
0
0
0
0
0
.
9
4
5
4
*
*
*
0
.
0
0
0
0
0
.
9
4
7
6
*
*
*
0
.
0
0
0
0
0
.
9
1
8
4
*
*
*
0
.
0
0
0
0
N
o
t
e
s
:
S
t
a
t
i
s
t
i
c
a
l
l
y
s
i
g
n
i
?
c
a
n
t
a
t
:
*
1
0
,
*
*
5
a
n
d
*
*
*
1
p
e
r
c
e
n
t
l
e
v
e
l
s
;
t
h
e
B
H
H
H
(
B
e
r
n
d
t
,
H
a
l
l
,
H
a
l
l
a
n
d
H
a
u
s
m
a
n
)
a
l
g
o
r
i
t
h
m
i
s
u
s
e
d
t
o
p
r
o
d
u
c
e
t
h
e
m
a
x
i
m
u
m
l
i
k
e
l
i
h
o
o
d
p
a
r
a
m
e
t
e
r
e
s
t
i
m
a
t
e
s
a
n
d
t
h
e
i
r
c
o
r
r
e
s
p
o
n
d
i
n
g
a
s
y
m
p
t
o
t
i
c
s
t
a
n
d
a
r
d
e
r
r
o
r
s
,
d
e
g
r
e
e
s
o
f
f
r
e
e
d
o
m
(
t
-
d
i
s
t
r
i
b
u
t
i
o
n
)
:
9
.
0
8
3
*
*
*
Table II.
Estimated coef?cient
of conditional
variance-covariance
equations, sample period:
April 1996-9 August 2007
JFEP
5,1
66
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
U
S
A
(
i
¼
1
)
E
M
U
(
i
¼
2
)
J
a
p
a
n
(
i
¼
3
)
C
h
i
n
a
(
i
¼
4
)
M
a
r
k
e
t
i
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
(
i
,
1
)
1
.
4
4
£
1
0
2
1
0
0
.
9
6
1
2
C
(
i
,
2
)
2
3
.
0
3
£
1
0
2
0
9
0
.
9
1
9
9
6
.
3
7
£
1
0
2
0
8
0
.
3
8
8
8
C
(
i
,
3
)
2
3
.
6
0
£
1
0
2
0
8
0
.
9
2
2
5
7
.
5
7
£
1
0
2
0
7
0
.
1
0
2
7
9
£
1
0
2
0
6
*
*
*
0
.
0
0
0
4
C
(
i
,
4
)
2
6
.
7
9
£
1
0
2
0
9
0
.
9
2
2
0
1
.
4
3
£
1
0
2
0
7
0
.
1
9
1
7
1
.
7
£
1
0
2
0
6
*
*
0
.
0
1
5
8
3
.
2
0
£
1
0
2
0
7
0
.
1
3
7
3
A
1
(
i
,
l
)
0
.
0
4
6
6
*
*
*
0
.
0
0
0
0
A
1
(
i
,
2
)
0
.
0
3
7
8
*
*
*
0
.
0
0
0
0
0
.
0
3
0
7
*
*
*
0
.
0
0
0
0
A
1
(
i
,
3
)
0
.
0
5
4
3
*
*
*
0
.
0
0
0
0
0
.
0
4
4
2
*
*
*
0
.
0
0
0
0
0
.
0
6
3
4
*
*
*
0
.
0
0
0
0
A
1
(
i
,
4
)
0
.
0
4
1
8
*
*
*
0
.
0
0
0
0
0
.
0
3
4
0
*
*
*
0
.
0
0
0
0
0
.
0
4
8
8
*
*
*
0
.
0
0
0
0
0
.
0
3
7
6
*
*
*
0
.
0
0
0
0
B
1
(
i
,
l
)
0
.
9
5
6
1
*
*
*
0
.
0
0
0
0
B
1
(
i
,
2
)
0
.
9
6
3
6
*
*
*
0
.
0
0
0
0
0
.
9
7
1
0
*
*
*
0
.
0
0
0
0
B
1
(
i
,
3
)
0
.
9
2
1
6
*
*
*
0
.
0
0
0
0
0
.
9
2
8
6
*
*
*
0
.
0
0
0
0
0
.
8
8
8
1
*
*
*
0
.
0
0
0
0
B
1
(
i
,
4
)
0
.
9
6
0
0
*
*
*
0
.
0
0
0
0
0
.
9
6
0
0
*
*
*
0
.
0
0
0
0
0
.
9
2
5
1
*
*
*
0
.
0
0
0
0
0
.
9
6
3
7
*
*
*
0
.
0
0
0
0
N
o
t
e
s
:
S
a
m
e
a
s
T
a
b
l
e
I
I
;
d
e
g
r
e
e
s
o
f
f
r
e
e
d
o
m
(
t
-
d
i
s
t
r
i
b
u
t
i
o
n
)
:
8
.
0
3
0
*
*
*
Table III.
Estimated coef?cient of
variance-covariance
equations, sample
period: 9 August,
2007-April 2011
Contagion
channels
67
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
While there are empirical evidence of contagion among EMU Japan and USA, the
Chinese market remained relatively immune after sub-prime crises.
Many factors worked favourably in keeping China’s equity market directly immune
from US turmoil. First, China has low exposure to US stock market. In contrary with
EMU and especially Japan, China preferred to buy USA’s dept via bonds, than invest in
the US economy. Furthermore, Chinese economy has fewer, relative to EMU and Japan,
bilateral trade and ?nancial relationships with USA.
Despite this fact, China exhibits strong spillover effects with Japan, which was
strongly affected from crisis[8]. The strong bilateral economic ties of Japan and China
explain the poor performance of Chinese equity market after the US subprime crisis
(Figure 1). Overall, contagion holds for all markets, but through different channels.
Table IV summarises testing hypothesis results. In all cases the null hypothesis of
no own or cross-spillover effects is rejected at 1 per cent signi?cance level. Moreover,
the reported Q-test statistics of Ljuing-Box, provide evidence of no serial
autocorrelation and therefore no misspeci?cation errors of the estimated diagonal
VECH model.
5. Conclusions
The impact of 2007 US subprime crisis on the equity markets of USA, EMU, China
and Japan are investigated, based on a VECH representation of multivariate
GARCH model. Empirical evidences indicate that the EMU and Japan have directly
impacted by the US subprime crisis, while China’s effects come through the Japanese
stock market. As a result, all markets exhibit contagion effects with the US market.
In line with several studies (Dooley and Hutchison, 2009; Kenourgios et al.,
2011; Dimitriou and Simos 2012) our empirical ?ndings support the conclusion that
an investment strategy ought to take into account all direct and indirect transmission
channels among markets during turmoil periods. A possible explanation for the
contagion effect in global markets is the “domino effect” created by investors.
Figure 2.
Cross-volatility spillover
effects before and
during the crisis
Note: Blue columns indicate before crisis period and green during crisis period
JFEP
5,1
68
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Since investors scramble to sell their assets and move into cash, produce higher
correlations and leading to contagion effects. Further research could consider the
asymmetric effects on conditional covariance through, for example, asymmetric
generalized dynamic conditional correlation models.
Notes
1. The revision of expectations theory suggests the existence of feedback traders where
asymmetric information could lead to the propagation of crises through portfolio
rebalancing effects.
2. Herding behavior emphasizes investors’ beliefs that asset prices contain relevant
information. Herding may encompass diverse phenomena such as bank runs, ?ckle
investors and hot money.
3. Other researchers, such as Chiang et al. (2007) and Kenourgios et al. (2011), employ a
multivariate GARCH-DCC speci?cation to estimate time-varying correlations (e.g. see Li et al.
(2012) for a survey on measurements of ?nancial interdependence, including contagion).
Pre-crisis period
First set of H
0
Wald test
H
0
a
1
¼ a
2
¼ a
3
¼ a
4
¼ 0 a
1
¼ 0 a
2
¼ 0 a
3
¼ 0 a
4
¼ 0
x
2
1,294.83
* * *
421.81
* * *
416.42
* * * *
313.62
* * *
426.92
* * *
p-value 0.0000 0.0000 0.0000 0.0000 0.0000
Deg. of freedom 4 1 1 1 1
Second set of H
0
H
0
b
1
¼ b
2
¼ b
3
¼ b
4
¼ 0 b
1
¼ 0 b
2
¼ 0 b
3
¼ 0 b
4
¼ 0
x
2
2,021,432.2
* * *
811,626.1
* * *
667,800.5
* * *
732,095.5
* * *
68,694.9
*
p-value 0.0000 0.0000 0.0000 0.0000 0.0000
Deg. of freedom 4 1 1 1 1
During crisis period
First set of H
0
Wald test
H
0
a
1
¼ a
2
¼ a
3
¼ a
4
¼ 0 a
1
¼ 0 a
2
¼ 0 a
3
¼ 0 a
4
¼ 0
x
2
424.77
* * *
221.95
* * *
257.51
* * *
86.33
* * *
164.55
* * *
p-value 0.0000 0.0000 0.0000 0.0000 0.0000
Deg. of freedom 4 1 1 1 1
Second set of H
0
H
0
b
1
¼ b
2
¼ b
3
¼ b
4
¼ 0 b
1
¼ 0 b
2
¼ 0 b
3
¼ 0 b
4
¼ 0
x
2
465,279.2
* * *
149,357.5
* * *
395,284.8
* * *
6,340.5
* * *
157.693.8
*
p-value 0.0000 0.0000 0.0000 0.0000 0.0000
Deg. of freedom 4 1 1 1 1
Pre-crisis period During crisis
period
Ljuing-Box test
H
0
: no serial
auto-correlation
Q-statistic 765.70 515.97
Lags 45 30
p-value 0.1156 0.1241
Note: Statistically signi?cant at:
*
10,
* *
5 and
* * *
1 per cent levels
Table IV.
Testing hypotheses of no
spillover effects and
Ljuing-Box test
Contagion
channels
69
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The VECH representation is chosen according to log-likelihood and AIC information criteria
(not reported here, but are available upon request).
4. According to Cecchetti (2009) and Baur (2012) the real trigger came on Thursday August 9,
the day that the large French bank BNP Paribas temporarily halted redemptions for three of
its funds that held assets backed by US subprime mortgage debt.
5. As suggested by Bekaert and Harvey (1995), calculating the returns in US dollars eliminates
the local in?ation.
6. The VAR order length is selected by the ?nal predicted error and the Akaike criterion. The
results are available upon request.
7. It is generally agreed that a GARCH (1,1) speci?cation with lag length one is adequate to
capture the dynamics of conditional variance (Bollerslev et al., 1998). Moreover, the AIC and
SIC criteria validate the GARCH (1,1) speci?cation.
8. Speci?cally, according to Nanto (2010), healthy ?nancial positions helped Mitsubishi UFG
Group, Japan’s largest bank, and Nomura, the country’s largest brokerage, to buy pieces of
distressed US investment banks as the crisis was deepening in October 2007. Mitsubishi
UFG bought 21 per cent of Morgan Stanley for 9 billion USD, and Nomura purchased the
Asian, European and Middle Eastern operations of Lehman Brothers. Similarly, EMU has
heavily exposure to USA’ equity markets via bilateral trade and investments.
References
Baur, D.G. (2012), “Financial contagion and the real economy”, Journal of Banking and Finance,
Vol. 36 No. 10, pp. 2680-98.
Bekaert, G. and Harvey, C.R. (1995), “Time-varying world market integration”, Journal of
Finance, Vol. 50, pp. 403-44.
Bekaert, G. and Harvey, C.R. (1997), “Emerging equity market volatility”, Journal of Financial
Economics, Vol. 43, pp. 29-77.
Bollerslev, T., Engle, R.F. and Wooldridge, J. (1988), “A capital asset pricing model with
time-varying covariances”, Journal of Political Economy, Vol. 96, pp. 116-31.
Brunnermeir, M. (2009), “Deciphering the liquidity and credit crunch 2007-2008”, American
Economic Association, Vol. 23, pp. 77-100.
Cecchetti, S. (2009), “Crisis and responses: the federal reserve in early stages of ?nancial
crisis”, Journal of American Perspectives, American Economic Association, Vol. 23,
pp. 51-75.
Chiang, T.C., Jeon, B.N. and Li, H. (2007), “Dynamic correlation analysis of ?nancial contagion:
evidence from Asian markets”, Journal of International Money and Finance, Vol. 26,
pp. 1206-28.
Dimitriou, D. and Simos, T. (2012), “International portfolio diversi?cation: an ICAPM approach
with currency risk”, Macroeconomics and Finance in Emerging Market Economies,
pp. 1-13.
Ding, Z. and Engle, R.F. (2001), “Large scale conditional covariance matrix modelling, estimation
and testing”, Academia Economics Papers, Vol. 29, pp. 157-84.
Dooley, M. and Hutchison, M. (2009), “Transmission of the USA subprime crisis to emerging
markets: evidence on the decoupling-recoupling hypothesis”, Journal of International
Money and Finance, Vol. 28, pp. 1331-49.
Forbes, K. and Rigobon, R. (2002), “No contagion, only interdependence: measuring stock market
comovements”, Journal of Finance, Vol. 57, pp. 2223-61.
JFEP
5,1
70
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Kaminsky, L. and Schukler, S. (1999), “What triggers market jitters? A chronicle of the Asian
crisis”, Journal of International Money and Finance, Vol. 18, pp. 537-60.
Kenourgios, D., Samitas, A. and Paltalidis, N. (2011), “Financial crises and stock market
contagion in a multivariate time-varying asymmetric framework”, International Financial
Markets, Institutions and Money, Vol. 21, pp. 92-106.
Kim, B.-H. and Kim, H. (2011), “Spillover effects of the US ?nancial crisis on ?nancial markets in
emerging Asian countries”, University of Auburn Working Paper Series, AUWP 2011-04.
King, M. and Wadhwani, S. (1990), “Transmition of volatility between stock markets”,
The Review of Financial Studies, Vol. 3, pp. 5-33.
Li, L., Zhang, N. and Willett, T.D. (2012), “Measuring macroeconomic and ?nancial market
interdependence: a critical survey”, Journal of Financial Economic Policy, Vol. 4, pp. 128-45.
Nanto, D.K. (2010), Global Financial Crisis, Analysis and Policy Implications, DIANE Publishing,
London.
Rose, A.K. and Spiegel, M.M. (2009), “Cross-country causes and consequences of the 2008 crisis:
international linkages and American exposure”, NBER Working Paper No. 1535822.
Saleem, K. (2008), “International linkage of Russian market and the Russian ?nancial crisis: a
multivariate GARCH analysis”, Discussion Papers 8/2008, Bank of Finland, BOFIT,
Institute for Economies in Transition.
Syllignakis, M. and Kouretas, G. (2011), “Dynamic correlation analysis of ?nancial contagion:
evidence from the Central and Eastern European markets”, International Review of
Economics and Finance, Vol. 20, pp. 717-32.
Worthington, A. and Higgs, H. (2004), “Transmission of equity returns and volatility in Asian
developed and emerging markets: a multivariate GARCH analysis”, International Journal
of Finance and Economics, Vol. 9, pp. 71-80.
Corresponding author
Dimitrios Dimitriou can be contacted at: [email protected]
Contagion
channels
71
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Dimitrios I. Dimitriou, Theodore M. Simos. 2014. Contagion effects on stock and FX markets. Studies
in Economics and Finance 31:3, 246-254. [Abstract] [Full Text] [PDF]
2. Dimitrios Dimitriou, Theodore Simos. 2013. Testing purchasing power parity for Japan and the US: A
structural-break approach. Japan and the World Economy 28, 53-59. [CrossRef]
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_417381589.pdf
The purpose of this paper is to investigate empirically contagion channels of the 2007 US
subprime financial crisis by employing a multivariate GARCH model for four major, international
equity markets, namely the USA, EMU, China and Japan.
Journal of Financial Economic Policy
Contagion channels of the USA subprime financial crisis: Evidence from USA, EMU,
China and Japan equity markets
Dimitrios Dimitriou Theodore Simos
Article information:
To cite this document:
Dimitrios Dimitriou Theodore Simos, (2013),"Contagion channels of the USA subprime financial crisis",
J ournal of Financial Economic Policy, Vol. 5 Iss 1 pp. 61 - 71
Permanent link to this document:http://dx.doi.org/10.1108/17576381311317781
Downloaded on: 24 January 2016, At: 21:46 (PT)
References: this document contains references to 21 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 288 times since 2013*
Users who downloaded this article also downloaded:
Salman Khan, Pierre Batteau, (2012),"Government intervention in Russian bourse: a case
of financial contagion", J ournal of Financial Economic Policy, Vol. 4 Iss 4 pp. 320-339 http://
dx.doi.org/10.1108/17576381211279299
Linyue Li, Nan Zhang, Thomas D. Willett, (2012),"Measuring macroeconomic and financial market
interdependence: a critical survey", J ournal of Financial Economic Policy, Vol. 4 Iss 2 pp. 128-145 http://
dx.doi.org/10.1108/17576381211228989
Simplice A. Asongu, (2012),"The 2011 J apanese earthquake, tsunami and nuclear crisis: Evidence of
contagion from international financial markets", J ournal of Financial Economic Policy, Vol. 4 Iss 4 pp.
340-353http://dx.doi.org/10.1108/17576381211279307
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Contagion channels of the USA
subprime ?nancial crisis
Evidence from USA, EMU, China and
Japan equity markets
Dimitrios Dimitriou and Theodore Simos
Department of Economics, University of Ioannina,
Ioannina, Greece
Abstract
Purpose – The purpose of this paper is to investigate empirically contagion channels of the 2007 US
subprime ?nancial crisis by employing a multivariate GARCH model for four major, international
equity markets, namely the USA, EMU, China and Japan.
Design/methodology/approach – In this study, contagion channels of the 2007 US subprime
?nancial crisis are investigated empirically by employing a multivariate GARCH model for four major,
international equity markets, namely the USA, EMU, China and Japan.
Findings – There is empirical evidence of contagion in all markets with the US market through
various channels, which have not been discussed in other related studies. Speci?cally, the empirical
results suggest that Japanese and EMU markets have been directly affected from the crisis. However,
while China’s equity market has been mainly unaffected by the US subprime crisis, has been affected
indirectly through Japan. Moreover, the Japanese equity market exhibits positive and signi?cant
spillovers effects with China and EMU, revealing an indirect volatility transmission channel of US
subprime crisis.
Research limitations/implications – Further research could consider the asymmetric effects
on conditional covariance through, for example, asymmetric generalized dynamic conditional
correlation models. All under examination markets show evidence of contagion through different
channels.
Practical implications – Despite the ?nancial advices for diversi?cation, since the increasing
globalization and stock market interdependence throughout the last 15 years, through the US subprime
crisis equity investors hadfewer opportunities for diversi?cation. Frompolicy makers’ perspective, they
should carefully examine and uncover possible decoupling strategies to insulate these economies from
contagion in future crises.
Social implications – This study provides useful information to international organizations, such
as World Bank and World Trade Organization (WTO) in order to protect markets from contagion
during future crises.
Originality/value – A novel ?nding of this paper is the indirect channel of contagion (i.e. Japanese
market) for Chinese market. This indirect channel may help explain why China’s equity market
performed badly in 2008 after the subprime crisis in the USA emerged.
Keywords United States of America, Japan, China, European Monetary Union, National economy,
Stock markets, Contagion channels, USA subprime crisis; M-GARCH models
Paper type Research paper
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JEL classi?cations – F3, F36, G01
We are grateful to two anonymous referees of this Journal for providing valuable comments
which substantially improved the quality of this article. Any remaining errors are ours.
Journal of Financial Economic Policy
Vol. 5 No. 1, 2013
pp. 61-71
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381311317781
Contagion
channels
61
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
1. Introduction
The 2007 US subprime ?nancial crisis, due to its severity and consequences to
international markets, is one of the most unanticipated economic events after the Great
Depression of 1930s and attracted great attention from academics, investors and policy
makers.
In this paper we empirically investigate contagion among the US market, the source
of the crisis, two major developed markets Japan and EMU and the most important
emerging market, China. The selection of these markets is based on market
capitalization and GDP in PPP terms. According to CIA World Factbook, at 1 January
2011 these markets are in the ?rst four positions according to the above criteria. Our
sample, weighted MSCI daily data, covers the period from April 1996 to April 2011,
including the 2007 US subprime crisis.
According to Forbes and Rigobon (2002) the contagion hypothesis is de?ned as
“a signi?cant increase in cross-market linkages after a shock to one country (or group
of countries)”, otherwise is considered to be “no contagion, only interdependence”.
Contagion is present during periods of serious worldwide turmoil. Therefore, the
measure of contagion and its transmission channels have been empirical issues under
intense investigation. Prior literature reported several crises transmission mechanisms.
King and Wadhwani (1990) and Kaminsky and Schukler (1999) suggested an analysis
based on revision of expectations[1] and herding[2] behaviour, respectively. In cases
of insulation during period of crises many researchers accept decoupling – recoupling
hypothesis, especially between developed and emerging markets (Dooley and
Hutchison, 2009).
The results from the literature investigating the contagion of US subprime crisis are
mixed. Dooley and Hutchison (2009) provide evidence of contagion on 14 emerging
markets. Kim and Kim (2011) ?nd evidence of ?nancial contagion, around the collapse
of Lehman Brothers in September 2008, on ?ve emerging Asian markets. Furthermore,
Syllignakis and Kouretas (2011) capture contagion effects among USA and German
stock market and seven emerging Central and Eastern Europe markets. On the other
hand, Rose and Spiegel (2009) ?nd little support of contagion using a wide number of
possible causes in a ?exible statistical framework for 107 countries.
Early research on the contagion effects used VAR and simple correlation analysis.
Our study applies a VECH representation of the multivariate GARCH (1,1) model to
estimate the spillover effects. Then we test their signi?cance during pre- and post-crisis
period. According to Ding and Engle (2001), a VECH representation is among the most
parsimonious versions of multivariate GARCH speci?cations[3].
This paper contributes the existent literature on several aspects. We empirically
investigate developed and emerging markets; by examining simultaneously the USA,
EMU, Japan and China stock market spillovers effects. In this way indirect channels of
contagion are revealed (i.e. one market is not affected directly by USA but indirectly
through another market).
Three interesting aspects emerged from our empirical analysis. First, we found
indirect channels of contagion for Chinese market via the Japanese market. As a result
China is indirectly affected by the US subprime crisis. Second, Japan and EMU markets
have been affected directly from the US crisis. Third, EMU directly affected by Japan.
The rest of the paper is structured as follows. In Section 2 we describe the
data employed and their descriptive statistics. The econometric techniques to estimate
JFEP
5,1
62
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the volatility spillovers are stated in Section 3. The empirical results and robustness
tests are discussed in Section 4. Last section concludes.
2. Data description and summary statistics
We employ a daily data set, comprising 3,912 observations for each market, and extend
over 15 years, spanning from April 1996 to April 2011. The data sample covers the US
?nancial crisis initiated in August 2007[4]. However, several other, more localized, crises
occurred during the sample period (e.g. Asian crisis in 1997, Russia’s rouble devaluation
in 1998, and the Brazilian real devaluation in 1999), although we focus mainly on the
recent US subprime due to the severity with which it gripped the markets and
economies globally. The returns are US-dollar denominated[5], computed using
weighted MSCI stock indices fromUSA, EMU, Japan and China equity markets. All data
are extracted from Datastream. The daily stock log-returns are evaluated using r
t
¼ ln
( p
t
/p
t21
) where p
t
is the market total index expressed in US dollars at time t.
Descriptive statistics for the pre-crisis and after-crisis periods are reported in Table I.
The return series appear to be non-normal due to excess kurtosis and fat tails,
supporting the chosen speci?cation for the GARCH innovations. Speci?cally, returns
from the Chinese market have a positive skewness, while returns from USA and EMU
are negatively skewed. Japan’s returns have positive skewness in pre-crisis period and
negative thereafter. The kurtosis, in all time series, exceeds three, indicating a
leptokurtic distribution. Excess kurtosis in equity returns has been well documented by
a number of other studies including Bekaert and Harvey (1997). Test results of the null
hypothesis of normality, using Jarque-Bera statistic, are reported in Table I. With all
USA EMU Japan China
Pre-crisis
Mean 8.72 £ 10
205
0.000167 1.07 £ 10
205
2.63 £ 10
205
Median 0.000410 0.000350 20.000161 20.000205
Maximum 0.053479 0.054348 0.109175 0.098866
Minimum 20.075337 20.057864 20.070541 20.130537
SD 0.010919 0.010002 0.013212 0.018652
Skewness 20.227170 20.222594 0.203966 0.044444
Kurtosis 6.6891 5.132132 6.225434 7.278809
Jarque-Bera 1,698.58
* * *
5,831.38
* * *
1,299.20
* * *
2,251.35
* * *
p-value (0.0000) (0.0000) (0.0000) (0.0000)
During crisis
Mean 26.20 £ 10
205
20.000319 20.000342 0.000181
Median 0.000514 0.000208 0.000000 0.000155
Maximum 0.109019 0.101241 0.104183 0.157124
Minimum 20.094087 20.098206 20.087861 20.135215
SD 0.017381 0.018807 0.017036 0.026602
Skewness 20.222718 20.035154 20.197545 0.234680
Kurtosis 9.54959 7.13294 7.64432 8.62253
Jarque-Bera 1,718.43
* * *
6,813.10
* * *
8,663.18
* * *
1,269.35
* * *
p-value (0.0000) (0.0000) (0.0000) (0.0000)
Notes: Statistically signi?cant at:
*
10,
* *
5 and
* * *
1 per cent levels; the Jarque-Bera LM statistic is
distributed asymptotically as x
2
(2) under the null hypothesis of normality
Table I.
Summary statistics
of daily returns of
international markets
Contagion
channels
63
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
p-values approximately zero to four decimal places, we reject the null hypothesis for all
markets. A visual perspective of the return’s volatility and the impact of 2007 crisis can
be obtained from graphs of stock indices levels, daily returns and conditional variances
for each market in Figure 1. The increase of volatility levels is noticeable in all markets
after the 2007 crisis, revealing clustering effects. Prior less sever crises partially explain
the increased volatility of stock returns during the period 1997-2000.
3. Econometric estimation of the spillover effects
We ?lter out the linear structure of the returns series to decouple them from the
conditional variance. To this end, we employ the VAR model:
r
t
¼ g þ
X
n
s¼1
a
s
r
t2s
þ1
t
; ð1Þ
where r
t
is the 4 £ 1 column vector of equity markets returns, g and a
s
are,
respectively, a 4 £ 1 vector and 4 £ 4 matrices of parameters and 1
t
are 4 £ 1 vectors of
innovations. The lag length, n ¼ 4, is chosen by information criteria[6]. A frequently
employed speci?cation of the conditional variance is Bollerslev et al. (1988)
representation of the multivariate GARCH model. In VECH form, the conditional
covariance matrix is given by:
vechðH
t
Þ ¼ c þ
X
q
j¼1
A
j
vechðu
t2i
u
0
t2i
Þ þ
X
p
j¼1
B
j
vechðH
t2j
Þ; ð2Þ
where vech( · ) denotes the column stacking operator of the lower portion of a symmetric
matrix. So, c is a N(N þ 1)/2 £ 1 vector and matrices A
j
and B
j
are of dimension
N(N þ 1)/2 £ N(N þ 1)/2.
The VECH representation of MGARCH is among the most parsimonious
multivariate GARCH speci?cations. Thus, we employ the diagonal VECH version of
multivariate GARCH (1,1) model[7] (Ding and Engle, 2001):
H
t
¼
CC
0
þ
AA
0
^u
t21
u
0
t21
þ
BB
0
^H
t21
; ð3Þ
Figure 1.
Markets returns,
conditional variance
and levels ?gures
JFEP
5,1
64
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
where u
i;t
jV
t21
, student 2ð0; H
t
; vÞ; V
t21
is the information set at time t 2 1, ^ is
the Hadamard (element by element) product. Since the null hypothesis of normality is
rejected, we assume that the conditional probability density function of the errors
follow the Student’s t-distribution. This is an empirically assumption that takes into
account the fat tail behaviour of the daily data set and, under correct speci?cation, may
improve the ef?ciency of the estimates. Since
CC
0
;
AA
0
and
BB
0
are all positive
semi-de?nite, H
t
will be positive de?nite for all t as far as the initial covariance matrix
H
0
is positive de?nite.
Let v
ij
¼ ð
C; C
0
Þ
ij
, a
ij
¼ ð
AA
0
Þ
ij
, b
ij
¼ ð
BB
0
Þ
ij
then:
h
ijt
¼ v
ij
þa
ij
u
it21
u
jt21
þb
ij
h
ijt21
; i; j ¼ 1; . . . ; N: ð4Þ
The elements of matrix A matrix measure the intensity of spillover effects among
markets. According to Chiang et al. (2007) when the cross-spillover coef?cients of
matrix A are positive, statistical signi?cant and rise after a crisis, we obtain evidence
of contagion. The elements of matrix B measure the persistence of conditional variance.
We estimate the rank one speci?cation of multivariate GARCH (1,1) model which is of
the form:
H
t
¼
CC
0
þ aa
0
^u
t21
u
0
t21
þ bb
0
^H
t21
; ð5Þ
where a and b are N £ 1 vectors. Here we impose the rank of parameter matrices A and
B to be one. The model is estimated using the full information maximum likelihood
(FIML) method with Student t-distributed errors. The FIML estimates are obtained by
maximizing the log-likelihood
P
T
t¼1
l
t
; where:
l
t
¼ log
Gððn þ mÞ=2Þn
m=2
ðnpÞ
m=2
Gðn=2Þðn 22Þ
m=2
2
1
2
logðjH
t
jÞ 2
1
2
ðn þ mÞlog 1 þ
1
0
t
H
21
t
1
t
n 22
" #
; ð6Þ
m is the number of equations, 1
t
is the m vector of residuals and v is the degree of
freedom.
4. Empirical results
At 9 August 2007 the BNP Paribas, a global investment bank, suspended its funds
after deteriorations in subprime mortgage loans. Several researchers used this date as
the starting point namely: Baur (2012), Brunnermeir (2019) and Cecchetti (2009) among
others. Based on this starting date we divide the sample into two sub-periods (pre- and
post-crisis). Estimates of linkages from equation (5), for each sub-period, are reported in
Tables II and III. In line with several other studies (Worthington and Higgs, 2004;
Saleem, 2008) cross-spillover estimates are positive and statistically signi?cant, for
both sample periods, indicating the presence of strong linkages. In Figure 2 we graph
the magnitude of cross-spillover effects for the two sub-periods. In the following
subsection we analyze these results extensively.
4.1 Interpretation of results and robustness tests
The estimated results of cross-spillover effects (Figure 2) indicate a signi?cant
increase, after crisis, for all pairs except of USA-China and EMU-China. These results
indicate direct and indirect channels of contagion for the markets under examination.
Contagion
channels
65
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
U
S
A
(
i
¼
1
)
E
M
U
(
i
¼
2
)
J
a
p
a
n
(
i
¼
3
)
C
h
i
n
a
(
i
¼
4
)
M
a
r
k
e
t
i
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
(
i
,
1
)
1
.
0
1
£
1
0
2
0
9
0
.
7
7
8
4
C
(
i
,
2
)
4
.
1
6
£
1
0
2
0
9
0
.
6
1
8
0
1
.
7
2
£
1
0
2
0
8
0
.
2
5
8
0
C
(
i
,
3
)
8
.
9
6
£
1
0
2
0
9
0
.
5
8
2
8
3
.
7
0
£
1
0
2
0
8
*
0
.
0
8
1
1
7
.
9
6
£
1
0
2
0
8
*
0
.
0
6
3
7
C
(
i
,
4
)
6
.
5
7
£
1
0
2
0
8
0
.
5
7
4
7
2
.
7
£
1
0
2
0
7
*
*
0
.
0
3
0
2
5
£
1
0
2
0
7
*
*
*
0
.
0
0
0
8
4
£
1
0
2
0
6
*
*
*
0
.
0
0
0
0
A
(
i
,
l
)
0
.
0
2
6
3
*
*
*
0
.
0
0
0
0
A
(
i
,
2
)
0
.
0
2
8
1
*
*
*
0
.
0
0
0
0
0
.
0
3
0
0
*
*
*
0
.
0
0
0
0
A
(
i
,
3
)
0
.
0
2
5
2
*
*
*
0
.
0
0
0
0
0
.
0
2
6
9
*
*
*
0
.
0
0
0
0
0
.
0
2
4
1
*
*
*
0
.
0
0
0
0
A
(
i
,
4
)
0
.
0
4
2
7
*
*
*
0
.
0
0
0
0
0
.
0
4
5
6
*
*
*
0
.
0
0
0
0
0
.
0
4
0
8
*
*
*
0
.
0
0
0
0
0
.
0
6
9
1
*
*
*
0
.
0
0
0
0
B
(
i
,
1
)
0
.
9
7
6
0
*
*
*
0
.
0
0
0
0
B
(
i
,
2
)
0
.
9
7
6
8
*
*
*
0
.
0
0
0
0
0
.
9
7
3
2
*
*
*
0
.
0
0
0
0
B
(
i
,
3
)
0
.
9
7
6
8
*
*
*
0
.
0
0
0
0
0
.
9
7
5
5
*
*
*
0
.
0
0
0
0
0
.
9
7
7
7
*
*
*
0
.
0
0
0
0
B
(
i
,
4
)
0
.
9
4
6
8
*
*
*
0
.
0
0
0
0
0
.
9
4
5
4
*
*
*
0
.
0
0
0
0
0
.
9
4
7
6
*
*
*
0
.
0
0
0
0
0
.
9
1
8
4
*
*
*
0
.
0
0
0
0
N
o
t
e
s
:
S
t
a
t
i
s
t
i
c
a
l
l
y
s
i
g
n
i
?
c
a
n
t
a
t
:
*
1
0
,
*
*
5
a
n
d
*
*
*
1
p
e
r
c
e
n
t
l
e
v
e
l
s
;
t
h
e
B
H
H
H
(
B
e
r
n
d
t
,
H
a
l
l
,
H
a
l
l
a
n
d
H
a
u
s
m
a
n
)
a
l
g
o
r
i
t
h
m
i
s
u
s
e
d
t
o
p
r
o
d
u
c
e
t
h
e
m
a
x
i
m
u
m
l
i
k
e
l
i
h
o
o
d
p
a
r
a
m
e
t
e
r
e
s
t
i
m
a
t
e
s
a
n
d
t
h
e
i
r
c
o
r
r
e
s
p
o
n
d
i
n
g
a
s
y
m
p
t
o
t
i
c
s
t
a
n
d
a
r
d
e
r
r
o
r
s
,
d
e
g
r
e
e
s
o
f
f
r
e
e
d
o
m
(
t
-
d
i
s
t
r
i
b
u
t
i
o
n
)
:
9
.
0
8
3
*
*
*
Table II.
Estimated coef?cient
of conditional
variance-covariance
equations, sample period:
April 1996-9 August 2007
JFEP
5,1
66
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
U
S
A
(
i
¼
1
)
E
M
U
(
i
¼
2
)
J
a
p
a
n
(
i
¼
3
)
C
h
i
n
a
(
i
¼
4
)
M
a
r
k
e
t
i
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
o
e
f
?
c
i
e
n
t
p
-
v
a
l
u
e
C
(
i
,
1
)
1
.
4
4
£
1
0
2
1
0
0
.
9
6
1
2
C
(
i
,
2
)
2
3
.
0
3
£
1
0
2
0
9
0
.
9
1
9
9
6
.
3
7
£
1
0
2
0
8
0
.
3
8
8
8
C
(
i
,
3
)
2
3
.
6
0
£
1
0
2
0
8
0
.
9
2
2
5
7
.
5
7
£
1
0
2
0
7
0
.
1
0
2
7
9
£
1
0
2
0
6
*
*
*
0
.
0
0
0
4
C
(
i
,
4
)
2
6
.
7
9
£
1
0
2
0
9
0
.
9
2
2
0
1
.
4
3
£
1
0
2
0
7
0
.
1
9
1
7
1
.
7
£
1
0
2
0
6
*
*
0
.
0
1
5
8
3
.
2
0
£
1
0
2
0
7
0
.
1
3
7
3
A
1
(
i
,
l
)
0
.
0
4
6
6
*
*
*
0
.
0
0
0
0
A
1
(
i
,
2
)
0
.
0
3
7
8
*
*
*
0
.
0
0
0
0
0
.
0
3
0
7
*
*
*
0
.
0
0
0
0
A
1
(
i
,
3
)
0
.
0
5
4
3
*
*
*
0
.
0
0
0
0
0
.
0
4
4
2
*
*
*
0
.
0
0
0
0
0
.
0
6
3
4
*
*
*
0
.
0
0
0
0
A
1
(
i
,
4
)
0
.
0
4
1
8
*
*
*
0
.
0
0
0
0
0
.
0
3
4
0
*
*
*
0
.
0
0
0
0
0
.
0
4
8
8
*
*
*
0
.
0
0
0
0
0
.
0
3
7
6
*
*
*
0
.
0
0
0
0
B
1
(
i
,
l
)
0
.
9
5
6
1
*
*
*
0
.
0
0
0
0
B
1
(
i
,
2
)
0
.
9
6
3
6
*
*
*
0
.
0
0
0
0
0
.
9
7
1
0
*
*
*
0
.
0
0
0
0
B
1
(
i
,
3
)
0
.
9
2
1
6
*
*
*
0
.
0
0
0
0
0
.
9
2
8
6
*
*
*
0
.
0
0
0
0
0
.
8
8
8
1
*
*
*
0
.
0
0
0
0
B
1
(
i
,
4
)
0
.
9
6
0
0
*
*
*
0
.
0
0
0
0
0
.
9
6
0
0
*
*
*
0
.
0
0
0
0
0
.
9
2
5
1
*
*
*
0
.
0
0
0
0
0
.
9
6
3
7
*
*
*
0
.
0
0
0
0
N
o
t
e
s
:
S
a
m
e
a
s
T
a
b
l
e
I
I
;
d
e
g
r
e
e
s
o
f
f
r
e
e
d
o
m
(
t
-
d
i
s
t
r
i
b
u
t
i
o
n
)
:
8
.
0
3
0
*
*
*
Table III.
Estimated coef?cient of
variance-covariance
equations, sample
period: 9 August,
2007-April 2011
Contagion
channels
67
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
While there are empirical evidence of contagion among EMU Japan and USA, the
Chinese market remained relatively immune after sub-prime crises.
Many factors worked favourably in keeping China’s equity market directly immune
from US turmoil. First, China has low exposure to US stock market. In contrary with
EMU and especially Japan, China preferred to buy USA’s dept via bonds, than invest in
the US economy. Furthermore, Chinese economy has fewer, relative to EMU and Japan,
bilateral trade and ?nancial relationships with USA.
Despite this fact, China exhibits strong spillover effects with Japan, which was
strongly affected from crisis[8]. The strong bilateral economic ties of Japan and China
explain the poor performance of Chinese equity market after the US subprime crisis
(Figure 1). Overall, contagion holds for all markets, but through different channels.
Table IV summarises testing hypothesis results. In all cases the null hypothesis of
no own or cross-spillover effects is rejected at 1 per cent signi?cance level. Moreover,
the reported Q-test statistics of Ljuing-Box, provide evidence of no serial
autocorrelation and therefore no misspeci?cation errors of the estimated diagonal
VECH model.
5. Conclusions
The impact of 2007 US subprime crisis on the equity markets of USA, EMU, China
and Japan are investigated, based on a VECH representation of multivariate
GARCH model. Empirical evidences indicate that the EMU and Japan have directly
impacted by the US subprime crisis, while China’s effects come through the Japanese
stock market. As a result, all markets exhibit contagion effects with the US market.
In line with several studies (Dooley and Hutchison, 2009; Kenourgios et al.,
2011; Dimitriou and Simos 2012) our empirical ?ndings support the conclusion that
an investment strategy ought to take into account all direct and indirect transmission
channels among markets during turmoil periods. A possible explanation for the
contagion effect in global markets is the “domino effect” created by investors.
Figure 2.
Cross-volatility spillover
effects before and
during the crisis
Note: Blue columns indicate before crisis period and green during crisis period
JFEP
5,1
68
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Since investors scramble to sell their assets and move into cash, produce higher
correlations and leading to contagion effects. Further research could consider the
asymmetric effects on conditional covariance through, for example, asymmetric
generalized dynamic conditional correlation models.
Notes
1. The revision of expectations theory suggests the existence of feedback traders where
asymmetric information could lead to the propagation of crises through portfolio
rebalancing effects.
2. Herding behavior emphasizes investors’ beliefs that asset prices contain relevant
information. Herding may encompass diverse phenomena such as bank runs, ?ckle
investors and hot money.
3. Other researchers, such as Chiang et al. (2007) and Kenourgios et al. (2011), employ a
multivariate GARCH-DCC speci?cation to estimate time-varying correlations (e.g. see Li et al.
(2012) for a survey on measurements of ?nancial interdependence, including contagion).
Pre-crisis period
First set of H
0
Wald test
H
0
a
1
¼ a
2
¼ a
3
¼ a
4
¼ 0 a
1
¼ 0 a
2
¼ 0 a
3
¼ 0 a
4
¼ 0
x
2
1,294.83
* * *
421.81
* * *
416.42
* * * *
313.62
* * *
426.92
* * *
p-value 0.0000 0.0000 0.0000 0.0000 0.0000
Deg. of freedom 4 1 1 1 1
Second set of H
0
H
0
b
1
¼ b
2
¼ b
3
¼ b
4
¼ 0 b
1
¼ 0 b
2
¼ 0 b
3
¼ 0 b
4
¼ 0
x
2
2,021,432.2
* * *
811,626.1
* * *
667,800.5
* * *
732,095.5
* * *
68,694.9
*
p-value 0.0000 0.0000 0.0000 0.0000 0.0000
Deg. of freedom 4 1 1 1 1
During crisis period
First set of H
0
Wald test
H
0
a
1
¼ a
2
¼ a
3
¼ a
4
¼ 0 a
1
¼ 0 a
2
¼ 0 a
3
¼ 0 a
4
¼ 0
x
2
424.77
* * *
221.95
* * *
257.51
* * *
86.33
* * *
164.55
* * *
p-value 0.0000 0.0000 0.0000 0.0000 0.0000
Deg. of freedom 4 1 1 1 1
Second set of H
0
H
0
b
1
¼ b
2
¼ b
3
¼ b
4
¼ 0 b
1
¼ 0 b
2
¼ 0 b
3
¼ 0 b
4
¼ 0
x
2
465,279.2
* * *
149,357.5
* * *
395,284.8
* * *
6,340.5
* * *
157.693.8
*
p-value 0.0000 0.0000 0.0000 0.0000 0.0000
Deg. of freedom 4 1 1 1 1
Pre-crisis period During crisis
period
Ljuing-Box test
H
0
: no serial
auto-correlation
Q-statistic 765.70 515.97
Lags 45 30
p-value 0.1156 0.1241
Note: Statistically signi?cant at:
*
10,
* *
5 and
* * *
1 per cent levels
Table IV.
Testing hypotheses of no
spillover effects and
Ljuing-Box test
Contagion
channels
69
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The VECH representation is chosen according to log-likelihood and AIC information criteria
(not reported here, but are available upon request).
4. According to Cecchetti (2009) and Baur (2012) the real trigger came on Thursday August 9,
the day that the large French bank BNP Paribas temporarily halted redemptions for three of
its funds that held assets backed by US subprime mortgage debt.
5. As suggested by Bekaert and Harvey (1995), calculating the returns in US dollars eliminates
the local in?ation.
6. The VAR order length is selected by the ?nal predicted error and the Akaike criterion. The
results are available upon request.
7. It is generally agreed that a GARCH (1,1) speci?cation with lag length one is adequate to
capture the dynamics of conditional variance (Bollerslev et al., 1998). Moreover, the AIC and
SIC criteria validate the GARCH (1,1) speci?cation.
8. Speci?cally, according to Nanto (2010), healthy ?nancial positions helped Mitsubishi UFG
Group, Japan’s largest bank, and Nomura, the country’s largest brokerage, to buy pieces of
distressed US investment banks as the crisis was deepening in October 2007. Mitsubishi
UFG bought 21 per cent of Morgan Stanley for 9 billion USD, and Nomura purchased the
Asian, European and Middle Eastern operations of Lehman Brothers. Similarly, EMU has
heavily exposure to USA’ equity markets via bilateral trade and investments.
References
Baur, D.G. (2012), “Financial contagion and the real economy”, Journal of Banking and Finance,
Vol. 36 No. 10, pp. 2680-98.
Bekaert, G. and Harvey, C.R. (1995), “Time-varying world market integration”, Journal of
Finance, Vol. 50, pp. 403-44.
Bekaert, G. and Harvey, C.R. (1997), “Emerging equity market volatility”, Journal of Financial
Economics, Vol. 43, pp. 29-77.
Bollerslev, T., Engle, R.F. and Wooldridge, J. (1988), “A capital asset pricing model with
time-varying covariances”, Journal of Political Economy, Vol. 96, pp. 116-31.
Brunnermeir, M. (2009), “Deciphering the liquidity and credit crunch 2007-2008”, American
Economic Association, Vol. 23, pp. 77-100.
Cecchetti, S. (2009), “Crisis and responses: the federal reserve in early stages of ?nancial
crisis”, Journal of American Perspectives, American Economic Association, Vol. 23,
pp. 51-75.
Chiang, T.C., Jeon, B.N. and Li, H. (2007), “Dynamic correlation analysis of ?nancial contagion:
evidence from Asian markets”, Journal of International Money and Finance, Vol. 26,
pp. 1206-28.
Dimitriou, D. and Simos, T. (2012), “International portfolio diversi?cation: an ICAPM approach
with currency risk”, Macroeconomics and Finance in Emerging Market Economies,
pp. 1-13.
Ding, Z. and Engle, R.F. (2001), “Large scale conditional covariance matrix modelling, estimation
and testing”, Academia Economics Papers, Vol. 29, pp. 157-84.
Dooley, M. and Hutchison, M. (2009), “Transmission of the USA subprime crisis to emerging
markets: evidence on the decoupling-recoupling hypothesis”, Journal of International
Money and Finance, Vol. 28, pp. 1331-49.
Forbes, K. and Rigobon, R. (2002), “No contagion, only interdependence: measuring stock market
comovements”, Journal of Finance, Vol. 57, pp. 2223-61.
JFEP
5,1
70
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Kaminsky, L. and Schukler, S. (1999), “What triggers market jitters? A chronicle of the Asian
crisis”, Journal of International Money and Finance, Vol. 18, pp. 537-60.
Kenourgios, D., Samitas, A. and Paltalidis, N. (2011), “Financial crises and stock market
contagion in a multivariate time-varying asymmetric framework”, International Financial
Markets, Institutions and Money, Vol. 21, pp. 92-106.
Kim, B.-H. and Kim, H. (2011), “Spillover effects of the US ?nancial crisis on ?nancial markets in
emerging Asian countries”, University of Auburn Working Paper Series, AUWP 2011-04.
King, M. and Wadhwani, S. (1990), “Transmition of volatility between stock markets”,
The Review of Financial Studies, Vol. 3, pp. 5-33.
Li, L., Zhang, N. and Willett, T.D. (2012), “Measuring macroeconomic and ?nancial market
interdependence: a critical survey”, Journal of Financial Economic Policy, Vol. 4, pp. 128-45.
Nanto, D.K. (2010), Global Financial Crisis, Analysis and Policy Implications, DIANE Publishing,
London.
Rose, A.K. and Spiegel, M.M. (2009), “Cross-country causes and consequences of the 2008 crisis:
international linkages and American exposure”, NBER Working Paper No. 1535822.
Saleem, K. (2008), “International linkage of Russian market and the Russian ?nancial crisis: a
multivariate GARCH analysis”, Discussion Papers 8/2008, Bank of Finland, BOFIT,
Institute for Economies in Transition.
Syllignakis, M. and Kouretas, G. (2011), “Dynamic correlation analysis of ?nancial contagion:
evidence from the Central and Eastern European markets”, International Review of
Economics and Finance, Vol. 20, pp. 717-32.
Worthington, A. and Higgs, H. (2004), “Transmission of equity returns and volatility in Asian
developed and emerging markets: a multivariate GARCH analysis”, International Journal
of Finance and Economics, Vol. 9, pp. 71-80.
Corresponding author
Dimitrios Dimitriou can be contacted at: [email protected]
Contagion
channels
71
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Dimitrios I. Dimitriou, Theodore M. Simos. 2014. Contagion effects on stock and FX markets. Studies
in Economics and Finance 31:3, 246-254. [Abstract] [Full Text] [PDF]
2. Dimitrios Dimitriou, Theodore Simos. 2013. Testing purchasing power parity for Japan and the US: A
structural-break approach. Japan and the World Economy 28, 53-59. [CrossRef]
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
6
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_417381589.pdf