Description
In September-October 2008 the Russian stock markets came under severe strain amidst
the global financial crisis. During this time the Russian government intervened several times to halt
the trade to impede the continuous slide. The government justified its actions owing to the argument
that the crisis was due to a trickledown effect from the financial crisis in the USA and the other
developed markets. The purpose of this paper is to put to test the government’s claim by exploring the
level of integration between Russia and the USA and European equity markets.
Journal of Financial Economic Policy
Government intervention in Russian bourse: a case of financial contagion
Salman Khan Pierre Batteau
Article information:
To cite this document:
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
Permanent link to this document:http://dx.doi.org/10.1108/17576381211279299
Downloaded on: 24 January 2016, At: 21:45 (PT)
References: this document contains references to 15 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 202 times since 2012*
Users who downloaded this article also downloaded:
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
Dimitrios Dimitriou, Theodore Simos, (2013),"Contagion channels of the USA subprime financial crisis:
Evidence from USA, EMU, China and J apan equity markets", J ournal of Financial Economic Policy, Vol. 5
Iss 1 pp. 61-71http://dx.doi.org/10.1108/17576381311317781
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
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Government intervention
in Russian bourse: a case
of ?nancial contagion
Salman Khan
Department of Finance, Suleman Dawood School of Business,
Lahore University of Management Sciences, Lahore, Pakistan, and
Pierre Batteau
Department of Finance, I.A.E Aix en Provence,
Universite´ Aix-Marseille II, Aix en Provence, France
Abstract
Purpose – In September-October 2008 the Russian stock markets came under severe strain amidst
the global ?nancial crisis. During this time the Russian government intervened several times to halt
the trade to impede the continuous slide. The government justi?ed its actions owing to the argument
that the crisis was due to a trickledown effect from the ?nancial crisis in the USA and the other
developed markets. The purpose of this paper is to put to test the government’s claim by exploring the
level of integration between Russia and the USA and European equity markets.
Design/methodology/approach – The study employs Markov Regime Switching Model for
tracking structural breaks in the time series. This method divides the data into three periods, i.e. pre
crisis, during crisis and post crisis. Next the Multivariate GARCH-DCC model technique is used to
establish the time varying linkages in order to verify the contagion effect. In the ?nal step the
Markowitz mean-variance framework is used to position each individual index portfolio with
respect to the ef?cient frontier to analyze the impact of crisis as well as Russian government
intervention.
Findings – The ?ndings suggest that the Russian equity market is weakly integrated with US and
strongly integrated with European markets. The results correspond to the underlying ?nancial and
economic linkages between Russia, the US and Europe. When examined in a portfolio setup, the results
show sudden fall in correlation among the Russian, US and European equity markets suggesting weak
linkages among these markets. Finally, the Markowitz Ef?cient frontier indicates dramatic rise in
volatility on the day intervention began and ended which signi?es the increased uncertainty among
the investors owing to Russian government ad hoc interventions.
Originality/value – The paper attempts to examine the Russian government intervention in the
backdrop of ?nancial crisis 2008 and concludes that the government intervention essentially increased
the uncertainty in the local as well as international markets. Therefore, it is essential that the
government should avoid direct intervention in its stock market.
Keywords Multivariate GARCH, Volatility spillovers, Financial crisis, Contagion, Stock markets,
Russia, United States of America, Europe
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?cation – C32, G15, G1
The authors wish to thank the Editor and two anonymous referees for constructive comments
on the earlier draft of this paper. The usual disclaimer applies.
JFEP
4,4
320
Journal of Financial Economic Policy
Vol. 4 No. 4, 2012
pp. 320-339
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381211279299
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
1. Introduction
The 15 September 2008 remains one of the most dramatic and memorable days in Wall
Street’s history when the Bank of America bought a near default Merrill Lynch for
$50 billion, Layman Brothers ?led bankruptcy and American International Group
(AIG) sought bailout from the US Government. On the 16 September 2008 the ?nancial
markets in the USA were in severe panic which triggered ?nancial crisis that
subsequently spread across the world markets.
For instance, the US, German, French, UK and Russian stock indices[1] recorded a
one day fall of 4.8, 2.8, 3.9, 4.0 and 5.0 percent, respectively, on 16 September 2008.
In the ensuing months, these stock markets continued to exhibit high levels of
volatility. Unlike US and European Governments, the Russian Government through
Federal Financial Market Service (FFMS) intervened in its stock market in order to
arrest the steep fall and suspended trade on 16 September 2009. The FFMS continued
to suspend the trade on different occasions spanning over 20 days; from 15 September
to 11 October 2008. The 20 days intervention appeared to be aligned with the steep fall
in US and European markets[2].
Throughout September-October 2008, the Russian Government claimed that the
shocks to Russian Trading System (RTS – a capital weighted index) directly resulted
from US subprime mortgage crisis and therefore necessitated government intervention.
The Russian Government held the viewthat trade -halt[3] should effectively isolate RTS
from external shocks and in turn provide time to investors for making informed
investment decisions (ef?cient market hypothesis-EMH). The Russian Government
adopted the strategy to halt the stock trade as soon as the markets in the USA and
Europe fell and it allowed the stock trade as soon as the US and European markets
exhibited the signs of recovery. Effectively, the government actions can be considered as
a hedging technique, i.e. to hedge the investors against market fall.
In this paper, we test the government claim that shocks from the ?nancial crisis
in 2008 transmitted from US and Europe to Russian stock markets, i.e. whether the
US and European stock markets can be considered contagious from Russian investor
perspective. In addition, we test the effectiveness of the government intervention in its
stock market in order to reduce the return-volatility and arrest the rapid decline
in stock prices.
At least in theory the notion of trade halt is acceptable since in the absence of trade the
price discovery[4] process does not take place. Such a theory would be more pronounced
in case of a single event such as the 9/11 terrorist attack on World Trade Center.
However, in the case of ?nancial crisis 2008, which spread over two months, this would
essentially mean that the government would halt the stock trade time and again
resultantly causing added uncertainty among the investors. In theory FFMS strategy
appears to be rewarding however, on the basis of risk-return, the Russian stock market
performed even worse than the US and European stock markets during the crisis.
2. Financial contagion
A market is regarded as ?nancially contagious if it spreads the shocks to other
markets. Edwards (2000) de?ne contagion as the transfer of shocks across countries
while Kaminsky and Reinhart (1999) de?ne contagion as the situation where the
knowledge of crisis in one country increases the risk of crisis in another country.
Forbes and Rigobon (2002) de?ne contagion as “a signi?cant increase in cross-market
Government
intervention in
Russian bourse
321
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
linkages after a shock to one country (or group of countries)” also called correlation
breakdown. Otherwise, a continued market correlation at high levels is considered to
be “no contagion, only interdependence”. When de?ning link between two or more
markets to be either contagious or interdependent is usually based on what triggers the
crisis in another market.
The most fundamental cause of contagion is considered to be a common shock,
for instance a major economic shift in industrial countries, a change in commodity prices,
a reduction in global growth, trade and ?nancial linkages ( Jokipii and Lucey, 2007).
Contagion can also be explained through the investor behavior theories such as the
liquidity problem. In this case investors cover their losses in one country by selling
securities inthe other markets inorder to raise cashinanticipationof greater redemptions.
Additionally, if banks experience amarkeddeteriorationinthe qualityof their loans toone
country, these banks may attempt to reduce the overall risk of their loan portfolio by
also minimizing their exposure in other high-risk investments, which could include other
emerging markets. Faced with liquidity problems, investors may be required to sell other
assets in their portfolios, ultimately leading to a fall in asset prices outside of the
crisis country, causing the disturbance to ripple through a variety of markets.
Much of the empirical work on measuring the existence of contagion is based on
comparing correlation coef?cients for interest rates, stock prices and sovereign spreads
between markets during a relatively stable period with a crisis or turbulent period
(King and Wadhwani, 1990; Boyer et al., 1999; Loretan and English, 2000; Forbes and
Rigobon, 2002; Corsetti et al., 2002).
We de?ne the contagion test given in Jokipii and Lucey (2007): if two markets
are naturally moderately correlated during periods of stability, then a shock to one
market will result in a signi?cant increase in market co-movements. This increase
constitutes contagion. If, on the other hand, the relationships do not change signi?cantly
after a shock to one market, and stability in the transmission mechanism persists, then
continued market co-movements can be inferred to as being driven by strong real
linkages betweenthe two economies. Suchstabilityinparameters over time woulddenote
interdependence. Based on these assumptions, contagion implies that cross-country
linkages are fundamentally different after a shock to one market while interdependence
implies no real change to relationships.
In establishing the relationship between the Russian and the foreign equity markets,
it is noteworthy that Gelos and Sahay (2000) and Saleem (2008) studied the 1998
Russian crisis and concluded in favor of contagion presence before the crisis in
the Russian market. Dungey et al. (2006) also found evidence of contagion from Russia
to international markets. These studies conclude that Russia is contagious from the
foreign markets perspective. More recently, Khan and Batteau (2011) evaluated the
performance of Russian Government intervention in RTS using event study
methodology. The study did not ?nd support in favor of government intervention.
However, the study is limited in its use of arbitrary linkages for cross-market volatility
transmission.
This paper attempt to ?ll this gap and attempt to answers following questions: does
Russia have ?nancial linkages with US and European stock markets? Are these
linkages strong enough to trigger contagion effect in the backdrop of ?nancial crisis
2008? Is there supporting evidence of underlying economic linkages which justi?es the
contagion effect?
JFEP
4,4
322
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In order to analyze the cross-country linkages most of the studies have focused on
multivariate volatility models. Consistent with these studies, we utilize the multivariate
generalized autoregressive conditional heteroscedasticity (GARCH) (dynamic
conditional correlation (DCC)) model. We establish DCC thresholds in order to clearly
understand and interpret the level of cross-market integration. Next we use Markowitz
ef?cient frontier framework to assess and verify the impact of Russian Government
intervention.
The rest of the paper is organized as follows: in Section 3 we highlight the economic
linkages underlying Russia and its counterparts, i.e. Europe and the USA. In Section 4
we discuss the data and preliminary analysis, in Section 5 we introduce GARCH-DCC
in bivariate and multivariate forms and the empirical results thereof, in Section 6 we
position each individual index portfolio with respect to the ef?cient frontier to analyze
the impact of crisis as well as Russian Government intervention and in Section 7 we
draw conclusions.
3. Economic linkages
The European Union (EU) and Russia have a strong trade relationship[5]. Bilateral
trade and investments continue to grow rapidly. Russia is one of the EU’s key trading
partners and trade between the two economies has showed steep growth rates until
mid-2008. This trend was interrupted by the economic crisis and the unilateral
measures adopted by Russia which have affected the bilateral trade[6]. The EU is by
far Russia’s main trading partner, accounting for 47.1 percent of its overall trade
turnover in 2010. It is also by far the most important investor in Russia. It is estimated
that up to 75 percent of foreign direct investment stocks in Russia come from the EU
Member States. EU good exports to Russia 2010 stands at $114.1 billion while EU
goods imports from Russia 2010 stands at $210.2 billion.
On the other hand, the level of bilateral ?nancial and economic exchange between
Russia and the USA is low[7]. Each only accounts for minor fractions of the other’s
trade with the rest of the world. Trade is dominated by inter-industry trade, with the
USA mainly exporting skill-intensive products to Russia while importing commodities.
This is not surprising given the countries’ different economic structures and factor
endowments. Russia accounted for only 0.7 percent of US exports and 1.3 percent of US
imports in 2008. Vice versa, the shares are slightly higher with the US accounting for
3.3 percent of Russia’s exports and 4.4 percent of Russia’s imports. Bilateral FDI shows
lower ?gures than trade: Russia accounted for only 0.3 percent ($6 billion) of the US
FDI stock abroad on average from 2000 to 2008. FDI increased sharply from 2003 to
2007 but recorded a signi?cant drop in 2008.
These economic linkages between Russia and Europe and the USA determine the
magnitude of shock transmitting from one market to another. Taking into account
these linkages, the correlation between Russia and Europe is expected to be stronger
and stable and the reverse is expected in case of Russia and the USA.
4. Data and preliminary analysis
Primary data consists of capital weighted indices including RTS 50 stock index[8]
(RTS), Standard and Poor’s 500 stock index (S&P500) – a US index, Deutscher Aktien
IndeX 30 (DAX30) – a German index, Cotation Assiste´e en Continu 40 stocks index
(CAC40) – a French index, Financial Times 100 stock index (FTSE100) – a UK index.
Government
intervention in
Russian bourse
323
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The data is obtained from DataStream comprising 2,266 daily observations from
3 April 2001 to 31 March 2010[9]. The closing prices of each index are converted to
natural log returns.
Apart from the underlying economic linkages as discussed in Section 3, the choice of
S&P500, DAX30, CAC40 and FTSE100 is due to the number of listed Russian
American Depository Receipts’ (ADR’s)[10] in these markets. Further support to
employ these series comes from the magnitude of trade speci?cally between Russia
and Germany, the UK, the USA and France. The area plot of the ?ve indices, i.e. RTS,
S&P500, DAX30, CAC40 and FTSE100 can be seen in Figure 1. The deep fall on
15 September 2008 has been attributed to the Lehman Brothers bankruptcy ?ling
among other events. The RTS appears more volatile from mid-2008 onwards.
As discussed earlier, the sudden fall on 15 September 2008 is quite steep compared to
the other stock indices. Since the time period under study includes a crisis period, we
need to identify appropriate breaks in the time series. This will allow us to analyze the
behavior of market correlation in different periods.
4.1 Structural breaks
Assume that y
t
is a time series generated as an autoregression of order r with regime
switching mean and variance:
y
t
¼ a
k
þb
k;S
t
x
k;t
þ1
t
ð1Þ
where:
1
t
, N 0; s
2
S
t
y
t
: RTS returns, a
k
: constant, b
k;S
t
: coef?cient for independent variables x
k,t
where
k ¼ 1, 2, 3, 4 representing S&P500, DAX30, CAC40 and FTSE100 returns, respectively;
1
t
: residual vector which follows a normal distribution; s
2
S
t
: the variance of the
Figure 1.
Daily closing prices of all
indices in the sample
7
Daily closing price
6
I
n
d
e
x
5
4
3
2
1
0
2002 2004 2006 2008 2010
RTS
S&P500
DAX
CAC40
FTSE100
JFEP
4,4
324
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
innovation at state S
t
and is assumed to be an n-state ?rst order Markov process,
taking the values 1, . . . ,n with transition probability matrix P ¼ P
ij
, i,j ¼ 1, . . . ,n
where:
P ¼ P
ij
½S
t
¼ jjS
t
¼ i? with
X
n
j¼1
P
ij
¼ 1 for all i:
Since S
t
is unobservable, the unknown parameters of the model can be estimated using
the non-linear ?lter proposed by Hamilton (1989). The model in equation (1) is
commonly known as Markov regime switching model (MRSM).
We estimate the model based on two states, i.e. low and high volatility for each pair
of index, i.e. RTS-S&P500, RTS-DAX30, RTS-CAC40 and RTS-FTSE100. The results
of MRSM are given in Table I.
For each pair of index, i.e. RTS-S&P500, RTS-DAX30, RTS-CAC40 and
RTS-FTSE100, we split the series into three periods based on two states:
(1) pre-crisis (3 April 2001-10 July 2008);
(2) crisis (11 July 2008-16 September 2009); and
(3) post-crisis (17 September 2009-31 March 2010).
Figure 2 shows the structural breaks in the sample markets. According to MRSM, the
crisis in RTS began on 11 July 2008, more than two months earlier than the beginning
of FFMS intervention. The FFMS intervened between 16 September 2008 and
11 October 2008 (20 days).
Figure 2 raises question regarding whether the timing of FFMS intervention
was consistent with the inception of crisis in RTS market. If the FFMS was of the view
that the RTS was suffering from the external shocks, then the FFMS should have
intervened in early July 2008. As a tertiary result, the Russian Government intervention
could be regarded as an ad hoc intervention.
4.2 Descriptive statistics
Table II gives the descriptive statistics for all the indices. These statistics has been
calculated using the daily closing returns of each index. The high return-volatility of
RTS in comparison to other indices is consistent with Harvey (1995) that emerging
stock markets show high volatility together with high returns. During pre-crisis,
the RTS offered high returns compared to the other stock markets while the riskiness
of RTS returns closely followed the other stock markets. The result shows that in
pre-crisis phase the RTS risk-return performance exceeded the other sample stock
markets.
During the crisis phase, all the markets exhibited negative returns along with high
levels of risks whereas RTS suffered the highest loss coupled with highest level of risk.
During this period, the risk-return of RTS was at extreme levels compared to the other
markets which points towards an abnormal activity that uniquely affected RTS
(possible the FFMS direct intervention). In post-crisis, RTS offered positive returns
with high risk compared to the other markets which essentially exhibited negative
returns with low level of risk. The argument shows the risk-return potential of RTS
market vis-a` -vis other markets.
Almost all the markets suffer from negative skewness and high kurtosis in all the
periods. During crisis the negative skewness shows that most of the returns lie under
Government
intervention in
Russian bourse
325
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
R
e
g
r
e
s
s
o
r
s
V
o
l
a
t
i
l
i
t
y
T
r
a
n
s
i
t
i
o
n
p
r
o
b
a
b
i
l
i
t
i
e
s
D
e
p
.
-
i
n
d
e
p
.
V
a
l
u
e
S
E
(
P
r
)
V
a
l
u
e
S
E
(
P
r
)
S
t
a
t
e
1
S
t
a
t
e
2
R
T
S
S
&
P
5
0
0
b
1
;
S
t
¼
1
0
.
3
3
*
0
.
0
4
(
0
.
0
0
)
s
21
;
t
0
.
0
0
0
2
*
0
.
0
0
(
0
.
0
0
)
p
1
1
¼
0
.
9
9
*
(
0
.
0
2
,
0
.
0
0
)
p
1
2
¼
0
.
0
6
*
(
0
.
0
1
,
0
.
0
0
)
b
1
;
S
t
¼
2
0
.
5
0
*
0
.
0
8
(
0
.
0
0
)
s
22
;
t
0
.
0
0
1
6
*
0
.
0
0
(
0
.
0
0
)
p
2
1
¼
0
.
0
1
*
(
0
.
0
0
,
0
.
0
0
)
p
2
2
¼
0
.
9
4
*
(
0
.
0
0
,
0
.
0
0
)
a
1
0
.
0
0
0
.
0
0
(
1
.
0
0
)
L
o
g
-
l
i
k
e
l
i
h
o
o
d
5
,
1
8
4
.
0
9
R
T
S
-
D
A
X
3
0
b
2
;
S
t
¼
1
0
.
3
4
*
0
.
0
2
(
0
.
0
0
)
s
21
;
t
0
.
0
0
0
2
*
0
.
0
0
(
0
.
0
0
)
p
1
1
¼
0
.
9
8
*
(
0
.
0
2
,
0
.
0
0
)
p
1
2
¼
0
.
0
6
*
(
0
.
0
2
,
0
.
0
0
)
b
2
;
S
t
¼
2
0
.
8
7
*
0
.
0
8
(
0
.
0
0
)
s
2
2
;
t
0
.
0
0
1
3
*
0
.
0
0
(
0
.
0
0
)
p
2
1
¼
0
.
0
2
*
(
0
.
0
0
,
0
.
0
0
)
p
2
2
¼
0
.
9
4
(
i
n
f
,
1
.
0
0
)
a
2
0
.
0
0
0
.
0
0
(
1
.
0
0
)
L
o
g
-
l
i
k
e
l
i
h
o
o
d
5
,
9
1
9
.
8
1
R
T
S
-
C
A
C
4
0
b
3
;
S
t
¼
1
0
.
4
0
*
0
.
0
3
(
0
.
0
0
)
s
21
;
t
0
.
0
0
0
2
*
0
.
0
0
(
0
.
0
0
)
p
1
1
¼
0
.
9
8
*
(
0
.
0
2
,
0
.
0
0
)
p
1
2
¼
0
.
0
6
*
(
0
.
0
1
,
0
.
0
0
)
b
3
;
S
t
¼
2
0
.
9
5
*
0
.
0
7
(
0
.
0
0
)
s
2
2
;
t
0
.
0
0
1
1
*
0
.
0
0
(
0
.
0
0
)
p
2
1
¼
0
.
0
2
*
(
0
.
0
0
,
0
.
0
0
)
p
2
2
¼
0
.
9
4
*
(
0
.
0
1
,
0
.
0
0
)
a
3
0
.
0
0
0
.
0
0
(
1
.
0
0
)
L
o
g
-
l
i
k
e
l
i
h
o
o
d
5
,
9
6
5
.
6
5
R
T
S
-
F
T
S
E
1
0
0
b
4
;
S
t
¼
1
0
.
5
1
*
0
.
0
3
(
0
.
0
0
)
s
21
;
t
0
.
0
0
0
2
*
0
.
0
0
(
0
.
0
0
)
p
1
1
¼
0
.
9
8
*
(
0
.
0
2
,
0
.
0
0
)
p
1
2
¼
0
.
0
5
*
(
0
.
0
1
,
0
.
0
0
)
b
4
;
S
t
¼
2
1
.
1
0
*
0
.
0
8
(
0
.
0
0
)
s
2
2
;
t
0
.
0
0
1
0
*
0
.
0
0
(
0
.
0
0
)
p
2
1
¼
0
.
0
2
*
(
0
.
0
0
,
0
.
0
0
)
p
2
2
¼
0
.
9
5
*
(
0
.
0
1
,
0
.
0
0
)
a
4
0
.
0
0
0
.
0
0
(
1
.
0
0
)
L
o
g
-
l
i
k
e
l
i
h
o
o
d
5
,
9
8
3
.
9
4
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
5
p
e
r
c
e
n
t
c
o
n
?
d
e
n
c
e
l
e
v
e
l
;
t
h
e
a
k
r
e
p
r
e
s
e
n
t
c
o
n
s
t
a
n
t
;
b
k
–
r
e
g
r
e
s
s
i
o
n
c
o
e
f
?
c
i
e
n
t
w
h
e
r
e
k
¼
1
(
S
&
P
5
0
0
)
,
2
(
D
A
X
3
0
)
,
3
(
C
A
C
4
0
)
a
n
d
4
(
F
T
S
E
1
0
0
)
;
s
2
1
;
t
–
v
a
r
i
a
n
c
e
o
f
t
h
e
i
n
n
o
v
a
t
i
o
n
a
t
s
t
a
t
e
S
t
w
h
e
r
e
S
t
¼
2
s
t
a
t
e
s
;
s
t
a
n
d
a
r
d
e
r
r
o
r
a
n
d
p
r
o
b
a
b
i
l
i
t
y
a
r
e
i
n
p
a
r
e
n
t
h
e
s
e
s
Table I.
Markov regime
switching model
JFEP
4,4
326
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the positive tail of the normal distribution indicating presence of extreme values
(the crisis) in left tail. The results from skewness and kurtosis are largely comparable
between RTS and the other stock markets which suggest RTS general behavior similar
to a stock market of a developed economy.
Figure 3 shows volatility clustering, i.e. large changes in returns are followed by
large changes while small changes are followed by small changes implying ARCH
effects. The strong ARCH effects can be found over the period ranging from July 2008
to September 2009 which is common to all the markets. This period is essentially
marked by the crisis phase. To deal with the ARCH effects, we use a parsimonious
model such as GARCH(1, 1)-DCC.
5. Dynamic conditional correlations
We use the following mean equation for estimation of each return series:
R
i;t
¼ a
i
þbR
i;t21
þ1
i;t
ð2Þ
where R
i,t
is the return on index i at time t, a
i
is a constant, b is the auto regression
coef?cient and 1
i,t
is the error term. The Engle (2002) GARCH-DCC model is a two-stage
estimator of conditional variances and correlations. In the ?rst stage, a univariate
GARCH model is estimated. In the second stage, the univariate variance estimates
obtained from the ?rst stage are used as inputs. The DCC model provides the time
Figure 2.
Estimated Markov regime
switching model
1
0.5
RTS-S&P500
RTS-DAX30
RTS-CAC40
RTS-FTSE100
0
1
0.5
0
1
0.5
S
m
o
o
t
h
P
r
o
b
a
b
i
l
i
t
i
e
s
0
1
0.5
0
2002 2004 2006 2008 2010
2002 2004 2006 2008 2010
2002 2004 2006 2008 2010
2002 2004 2006 2008 2010
State 1 State 2 Intervention begins Intervention ends
Government
intervention in
Russian bourse
327
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 3.
Daily return volatility of
all indices in the sample
0.4
–0.2
0.2
0.3
0.2
0.1
0
–0.1
0.3
0.2
0.1
0
–0.1
0
–0.4
0.3
0.2
0
0.1
2000
DAX
FTSE100 CAC40
S&P500
RTS
2005
2005 2000
2010 2000 2005 2010
2000 2005 2010 2000 2005 2010
2010
0.05
0
–0.05
–0.1
–0.1
0.1
Complete Pre-crisis Crisis Post-crisis
RTS Mean ( £ 1,000) 0.99 1.40 21.92 0.94
Median ( £ 1,000) 2.38 2.45 0.45 2.43
SD (%) 2.27 1.74 4.35 2.29
Skewness 20.49 20.54 20.16 20.51
Kurtosis 14.10 6.22 6.89 14.32
S&P500 Mean ( £ 1,000) 0.01 0.05 20.54 20.03
Median ( £ 1,000) 0.59 0.47 1.33 0.52
SD (%) 1.37 1.06 2.64 1.39
Skewness 20.10 0.09 20.08 20.08
Kurtosis 11.94 5.54 5.75 11.88
DAX30 Mean ( £ 1,000) 0.03 0.05 20.32 20.00
Median ( £ 1,000) 0.79 0.79 0.59 0.77
SD (%) 1.68 1.55 2.43 1.70
Skewness 0.06 20.12 0.37 0.06
Kurtosis 7.43 6.14 6.48 7.39
CAC40 Mean ( £ 1,000) 20.12 20.11 20.33 20.15
Median ( £ 1,000) 0.25 0.24 0.99 0.26
SD (%) 1.59 1.41 2.51 1.61
Skewness 0.06 20.05 0.20 0.07
Kurtosis 8.32 6.39 6.22 8.36
FTSE100 Mean ( £ 1,000) 0.00 20.02 20.15 20.04
Median ( £ 1,000) 0.47 0.39 0.47 0.39
SD (%) 1.34 1.15 2.27 1.36
Skewness 20.11 20.16 20.02 20.10
Kurtosis 9.80 6.42 6.48 9.81
Table II.
Descriptive statistics
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
varying correlations among the variables that indicates whether the co-movement of
the variables increase/decrease over time. This modeling technique is useful in
analyzing the level of market integration, a condition prerequisite for effective shock
transmission. Following Engle (2002), the DCC-GARCH model can be formulated
as follows:
r
t
jV
ðt21Þ
, Nð0; D
t
R
t
D
t
Þ ð3Þ
D
2
t
¼ diag{v
i
} þ diag{k
i
}r
t21
+r
t21
þ diag{l
i
}D
2
t21
ð4Þ
1
t
¼ D
21
t
r
t
ð5Þ
Q
t
¼ Sð1 2a 2bÞ þa 1
t21
1
0
t21
À Á
þbQ
t21
ð6Þ
R
t
¼ diag{Q
t
}
21
Q
t21
diag{Q
t
}
21
ð7Þ
where equation (5) represents the standardized errors, S is the unconditional
correlation matrix of the errors and W is the Hadamard product of two matrices of the
same size. The parameters of the DCC-GARCH model can be estimated using
maximum likelihood. If a þ b , 1 then equation (6) is mean reverting and the
log-likelihood for this estimator can be written as:
L ¼ 21=2
X
T
ðt¼1Þ
ðnlogð2pÞ þ 2 logðjD
t
jÞ þ logðjR
t
jÞ þ1
0
t
R
21
t
1
t
Þ
where D
t
¼ diag
??????
h
i;t
p È É
and R
t
is used to construct time varying correlations. We also
establish the rules for interpreting correlation between two or more series, i.e. strong
correlation: between 0.60 and 0.99, moderate correlation: between 0.40 and 0.59, weak
correlation: between 0.10 and 0.39.
5.1 Bivariate dynamic correlation
We estimate the GARCH(1, 1)-DCC model as laid down in equations (1)-(7). The results
are illustrated in Table III. The ARCH coef?cients a
RTS
and a
k
and the GARCH
coef?cients b
RTS
and b
k
are found to be signi?cant at the 5 percent con?dence level,
where k ¼ S&P500, DAX30, CAC40, FTSE100. The constant v
RTS
and v
k
are also
signi?cant at the 5 percent con?dence level. All the series show considerable level of
ARCH and GARCH effects.
The coef?cient a
m
is signi?cant at the 5 percent level, indicating the lingering effect
of standardized residuals in the previous period. The coef?cient b
n
is also signi?cant
at the 5 percent level, indicating the memory of correlations. For the entire sample
(a
m
þ b
n
) is close to 1, implying that the volatility is exhibiting a highly persistent
behavior. The high level of shocks (a
RTS
) originating in RTS indicate that RTS has
signi?cant role in in?uencing the correlation between RTS and the US and European
markets. On the other hand, the volatility spillover or GARCH levels (b
RTS
) are higher
in case of foreign markets compared with the RTS implying that the volatility spillover
in US and European markets largely affects correlation relationship with RTS. These
results signify that the RTS was less prone to the shocks and more to the volatility
spillover effects from the foreign equity markets. We plot the dynamic correlations
from Q
t
in Figure 4[11].
Government
intervention in
Russian bourse
329
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
All the paired correlations behave almost in a similar fashion. These begin sliding a
month before the crisis began, followed by a steep fall and then start rising.
The moderate correlation (#0.5) indicates that any setback in US or European
market does not have the potential to create a wider panic in RTS market and vice
versa, a result that supports the low level ARCH effects associated with these markets.
During the crisis, the correlation dips for a few months begin to move upwards during
the intervention period. The rectangular boxes in Figure 4 provide contagion build
RTS-S&P500 RTS-DAX30 RTS-CAC40 RTS-FTSE100
Value SE Value SE Value SE Value SE
v
RTS
0.12
*
0.00 0.12
*
0.00 0.12
*
0.00 0.12
*
0.00
a
RTS
0.10
*
0.00 0.11
*
0.00 0.11
*
0.00 0.11
*
0.00
b
RTS
0.87
*
0.00 0.86
*
0.00 0.86
*
0.00 0.86
*
0.00
v
k
0.01
*
0.00 0.02
*
0.00 0.02
*
0.00 0.01
*
0.00
a
k
0.07
*
0.00 0.09
*
0.00 0.09
*
0.00 0.09
*
0.00
b
k
0.93
*
0.00 0.90
*
0.00 0.90
*
0.00 0.90
*
0.00
a
m
0.01
*
0.00 0.02
*
0.00 0.02
*
0.00 0.02
*
0.00
b
n
0.99
*
0.00 0.97
*
0.00 0.97
*
0.00 0.97
*
0.00
Log
a
27,796.41 28,280.39 28,105.31 27,600.08
LBQ(i ) 19.01 (0.27)
*
17.45 (0.36)
*
18.89 (0.27)
*
16.24 (0.44)
*
LBQ(i )
2
16.53 (0.42)
*
13.46 (0.64)
*
13.84 (0.61)
*
16.16 (0.44)
*
LBQ(k) 31.24 (0.01) 18.96 (0.27)
*
29.36 (0.02) 25.67 (0.06)
*
LBQ(k)
2
18.78 (0.28)
*
166.15 (0.0) 230.34 (0.0) 545.46 (0.0)
Notes: Parameters signi?cant at:
*
5 percent con?dence level;
a
log likelihood function; LBQ – Ljung
Box Q test performed at 16th lag; LBQ(i ) – where i ¼ RTS; LBQ(k) – where k ¼ S&P500, DAX30,
CAC40, FTSE100
Table III.
GARCH-DCC (1, 1)
paired estimates
Figure 4.
Bivariate dynamic
conditional
correlation-GARCH(1,1)
1
0.5
RTS-S&P500
RTS-DAX30
RTS-CAC40
RTS-FTSE100
0
1
0.5
0
1
0.5
0
1
0.5
0
Contagion buildup
Contagion buildup
Contagion buildup
Contagion buildup
Contagion
Contagion
Contagion
Contagion
2008 2010 2009
2008 2010 2009
2008 2010 2009
2008 2010 2009
S
m
o
o
t
h
P
r
o
b
a
b
i
l
i
t
i
e
s
(
M
R
S
M
)
/
C
o
r
r
e
l
a
t
i
o
n
s
Intervention end Intervention begin Pre crisis crisis Post crisis Dynamic correlations
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
up phases and trigger the contagion process. The slow decay in correlations at the
beginning of the crisis indicates that the investors in foreign markets were liquidating
their positions. Once the views of the investors in RTS and foreign markets converge,
the correlations increases dramatically, i.e. contagion takes place. From government
intervention perspective, it is clear that halting the trade could not isolate RTS from the
global markets linkages. We see in Figure 4 that the bivariate correlation continues to
increase instead of decrease due to intervention. This provides evidence that the
halting the trade was not the solution even if the other markets are deemed as
contagious.
The descriptive statistics of correlations for the entire sample is given in Table IV.
The mean, maximum and standard deviation in correlations are much lower for
RTS-S&P500 (weakly correlated) compared to other paired indices indicating low level
of shock transmission and volatility spillover between these markets. In case of
RTS-DAX30, RTS-CAC40 and RTS-FTSE100, the correlations are moderate and suffer
from high volatility in pre-crisis, low volatility during crisis and post-crisis periods as
indicated by the standard deviation in Table IV. Here it is noteworthy that the Russian
Government pointing towards the USA as a contagious market is out of place since
the RTS correlation statistics are lower.
The reasons that explain as why RTS appears to be more correlated with the
European stock markets compared to US stock market lies in the underlying ?nancial
and economic linkages between these countries. As discussed earlier, Russia has
strong ?nancial and economic linkages with the Europe compared to the USA.
So far our results suggest that all the markets in the sample may be regarded as
contagious and the Russian Government intervention was not successful in reducing
the volatility in RTS market as well. To add further support to these interim ?ndings,
we conduct analysis by constructing portfolios that includes all sample indices with
and without RTS. If the inclusion of RTS increases the overall correlation of the
Parameters Pre-crisis Crisis Post-crisis
RTS-S&P500 Mean 0.22 0.30 0.43
Max. 0.37 0.39 0.49
Min. 0.03 0.11 0.35
SD 0.07 0.06 0.04
RTS-DAX30 Mean 0.35 0.47 0.56
Max. 0.67 0.61 0.65
Min. 20.02 0.19 0.44
SD 0.13 0.09 0.05
RTS-CAC40 Mean 0.39 0.50 0.61
Max. 0.68 0.66 0.68
Min. 20.01 0.18 0.48
SD 0.14 0.10 0.05
RTS-FTSE100 Mean 0.40 0.51 0.61
Max. 0.68 0.67 0.69
Min. 0.04 0.22 0.48
SD 0.13 0.09 0.05
Notes: Strong correlation: between 0.60 and 0.99; moderate correlation: between 0.40 and 0.59; weak
correlation: between 0.10 and 0.39
Table IV.
Descriptive statistics
of correlations
Government
intervention in
Russian bourse
331
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
portfolio then the Russian Government’s claim of shocks pouring from foreign markets
during crisis can be well endorsed. On the other hand if the inclusion of RTS decreases
the correlation among the global markets then the government claim could be
fairly rejected.
5.2 DCC-portfolio perspective
We estimate the GARCH(1, 1)-DCC model given in equations (1)-(7) for two different
group of portfolios:
(1) Portfolio-I consists of S&P500, DAX30, CAC40, FTSE100; and
(2) Portfolio-II consists of S&P500, DAX30, CAC40, FTSE100 and RTS.
For the ?rst portfolio, the results are given in Table V, Panel (a). The ARCH coef?cients
a
a
and the GARCH coef?cients b
a
are found to be signi?cant at the 5 percent
con?dence level. The shocks (a
a
) for DAX30, CAC40, and FTSE100 are relatively
higher than the shocks for S&P500 which indicate that the shocks originated in these
markets affects the USA more while the USA affects the European markets in terms of
volatility spillover (b
a
) effects. With regard to the persistence, the results for both
Panels (a) and (b) indicate that the sum of the estimated coef?cients of the variance
equation (a
m
þ b
n
) is close to unity. This implies that volatility exhibits a highly
persistent behavior in both groups.
For the second portfolio, the results are given in Table V, Panel (b). The
ARCHcoef?cients shocks (a
b
) and the volatility spillover (b
b
) are found to be signi?cant
at the 5 percent con?dence level. The shocks (a
b
) for RTS are high compared to other
markets which indicate that the shocks emanating from RTS are transmitted to these
markets. On the other hand, the volatility spillover (b
b
) is higher for the US and
RTS S&P500 DAX30 CAC40 FTSE100
Value SE Value SE Value SE Value SE Value SE
Panel (a): S&P500, DAX30, CAC40 and FTSE100
v
a
2.89
*
0.86 3.56
*
0.98 3.21
*
0.86 1.51
*
0.22
a
a
0.12
*
0.00 0.15
*
0.00 0.14
*
0.00 0.13
*
0.00
b
a
0.86
*
0.00 0.85
*
0.00 0.85
*
0.00 0.86
*
0.00
a
m
0.04
*
0.00 Log likelihood 230,641.45
b
n
0.94
*
0.00
LBQ(a) 40.40 (0.00) 28.04 (0.03) 25.85 (0.06)
*
17.13 (0.38)
*
LBQ(a)
2
16.66 (0.41)
*
10.81 (0.82)
*
10.99 (0.81)
*
14.16 (0.59)
*
Panel (b): RTS, S&P500, DAX30, CAC40 and FTSE100
v
b
0.11
*
0.00 0.01
*
0.00 0.02
*
0.00 0.01
*
0.00 0.01
*
0.00
a
b
0.10
*
0.00 0.07
*
0.00 0.09
*
0.00 0.08
*
0.00 0.09
*
0.00
b
b
0.87
*
0.00 0.92
*
0.00 0.91
*
0.00 0.92
*
0.00 0.91
*
0.00
a
m
0.02
*
0.00 Log likelihood 214,086.06
b
n
0.97
*
0.00
LBQ(b) 10.22 (0.85)
*
43.59 (0.00) 27.00 (0.04) 25.37 (0.06)
*
17.99 (0.32)
*
LBQ(b)
2
12.55 (0.71)
*
26.00 (0.05)
*
32.74 (0.01) 6.99 (0.97)
*
19.22 (0.26)
*
Notes: Parameters signi?cant at:
*
5 percent con?dence level; LBQ – Ljung Box Q test performed at
16th lag where a ¼ S&P500, DAX30, CAC40, FTSE100 and b ¼ RTS, S&P500, DAX30, CAC40,
FTSE100
Table V.
GARCH-DCC (1, 1)
group estimates
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
European markets compared to RTS. This suggest that the US and European markets
volatility spillover can upset the correlation with the RTS market, a result similar to
bivariate GARCH(1, 1)-DCC analysis. The comparative conditional correlation for the
two groups of samples is plotted in Figure 5 within MRSM setup.
The correlation for the Portfolio-I (i.e. excluding RTS) is strong with a mean greater
than 0.6 in all periods compared to the Portfolio-II (i.e. including RTS) where the mean
correlations are found to be less than 0.45. The descriptive statistics for the correlations
are given in Table VI. The high correlations in the Portfolio-I over all periods indicate
high level of market integration implying strong shock transmission among the
markets. Portfolio-II reveals low level of correlations and therefore implies low level
of shock transmission.
The US-Europe certainly enjoys a higher level of correlation (less volatile) which
indicates that the US and European markets are interdependent. As illustrated in
Table VI, the mean correlation reduces from 0.60 to 0.23 as RTS is included in the
Portfolio-II. This means that RTS changes the dynamics of relationship between US and
Figure 5.
Dynamic conditional
correlation-GARCH(1,1)-
portfolio setup
0.9
1
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
2009 2008 2010
Dynamic Correlation Comparison (Two Portfollos)
Interdependence
Contagion
Contagion Buildup
Pre crisis crisis Post crisis Portfolio (excI.RTS)
Intervention begin Intervention end Portfolio (incl.RTS)
Pre-crisis Crisis Post-crisis
Portfolio-I (excl. RTS) Mean 0.60 0.68 0.69
Max. 0.78 0.77 0.79
Min. 0.24 0.51 0.53
SD 0.09 0.06 0.06
Portfolio-II (incl. RTS) Mean 0.23 0.31 0.45
Max. 0.52 0.45 0.55
Min. 20.02 20.00 0.33
SD 0.09 0.09 0.07
Notes: Strong correlation: between 0.60 and 0.99; moderate correlation: between 0.40 and 0.59; weak
correlation: between 0.10 and 0.39
Table VI.
Descriptive statistics
of correlations
Government
intervention in
Russian bourse
333
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
European equity markets. The dramatic fall in correlation and rise in volatility
(in correlation) when switching from Portfolio-I to II suggests that the inclusion of RTS
contributes towards the overall risk of the portfolio.
The rise in correlation in Portfolio-II is quite subtle during the intervention period
compared to Portfolio-I. One can argue that the continuous rise during the intervention
period can be attributed to the Russian Government intervention, however, one need to
be cautious in interpreting this particular argument since we do not know what the
correlations would have been in the absence of government intervention.
The evidence suggest that RTS is weakly correlated with the global markets in
a portfolio setup implying that simultaneous shocks from these markets does not have
the potential to create wide spread crisis in RTS. However, the RTS does have the
potential to spillover volatility to these markets. This essentially means that RTS can be
regarded as a contagious market. In addition, similar argument can be used to justify
the portfolio diversi?cation bene?ts once the RTS is included in the portfolio. In the next
section we study the Portfolio-II in an ef?cient frontier framework with conditional risk
element. The analysis will allow us to understand the in?uence of each stock index
on the ef?cient frontier on crisis and intervention time periods.
6. Portfolio implication from intervention
In this section we evaluate the effects of intervention using portfolio theory. It would be
interesting to know the ef?cient frontiers and the performance of RTS vis-a` -vis other
indices including the respective position of each stock index. We construct ef?cient
frontier using mean-variance framework. We reutilize the conditional covariance
matrix estimated in the previous section for Portfolio-II which includes all the sample
indices and is shown in Figure 6.
The graph shows the time evolution of ef?cient frontier. It is quite vivid that in
2008-2009 all the world markets were going through the ?nancial crisis. During the
?nancial crisis the returns on ef?cient portfolios were as high as 50 percent per annum
and as low as 225 percent per annum. Similarly the risk ranged from a low level of
2 percent per annum to as high as 20 percent per annum. These re?ects the abnormal
Figure 6.
Mean-variance ef?cient
frontier (risk-return in
annualized percentage)
60
50
40
30
20
10
–10
–20
2000
2002
2004
2006
2008
2010
2012
Time Evolution of Efficient Frontier
0
5
10
15
20
25
Portfolio Risk
P
o
r
t
f
o
l
i
o
R
e
t
u
r
n
s
50
40
30
20
10
0
–10
0
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
times where even the ef?cient portfolios based on well diversi?ed sub portfolio such as
the sample indices are offering freighting risk-return. The subtleties of individual
index risk-return cannot be adjudged through a three dimensional ?gure. Therefore, it
is necessary that we analyze the performance of different indices in the portfolio at
different point of time using dynamic correlations and conditional volatilities. For this
purpose, we select four days in 2008-2009, i.e. the day the crisis begin, the days when
intervention begins and ends as well as on the day the crisis ends. Figure 7 shows these
four ef?cient frontiers. The risk in these ?gures re?ects the conditional variances
obtained from GARCH-DCC Portfolio-II estimation.
One can easily spot the behavior of RTS compared to other indices. RTS which
appear to be close to other indices on the day the crisis begin shored away from other
indices. Also, it is quite evident that RTS extreme position on the selected dates
in?uenced the ef?cient frontier to a great extent. Let us look at the individual index
positions in more detail.
As the crisis begins (11 July 2008) the RTS index portfolio return is better (1 percent
per annum) than the other indices except S&P500 (2 percent per annum). The positive
return on RTS comes with greater level of risk (12 percent per annum) compared
to other. On the intervention day (16 September 2008), the RTS return deteriorated to
230 percent per annum. The RTS risk pro?le dramatically changed as well on the
intervention day, i.e. from 12 percent per annum to 130 percent per annum. The risk
and return of the other index portfolios doubled from the crisis level, while witnessing
fall in returns except for S&P500 where the returns improved.
At the end of the intervention (11 October 2008), when the RTS was expected to
demonstrate that the government actions actually caused considerable reduction in the
risk, instead we ?nd that the risk increased to 275 percent per annum nevertheless the
returns improved by huge margin, i.e. 24 percent per annum compared to other indices.
This behavior of RTS appears consistent with the portfolio theory that high risk yields
Figure 7.
Mean-variance
ef?cient frontier
Government
intervention in
Russian bourse
335
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
high return. In this period, the other markets also suffered fromhigh volatility compared
to RTS and the returns in these markets improved as well. The large risk-gap between
RTS and the other index portfolios provide evidence that the intervention did not had
substantial effect on stabilizing the market conditions in RTS. The evidence con?rms to
earlier results that intervention created additional risk in RTS.
When the crisis ended on 17 September 2009, the RTS risk came down to 23 percent
per annum while the returns came down to 4.6 percent per annum. Table VII provides
key investment performance measures of these portfolios.
The tracking error is small for all the indices compared to RTS. The Jensen’s a is
negative for all the portfolios in all the periods’ except post-crisis day indicating
absence of positive alpha pro?ts for holding index portfolios. The M
2
a is positive and
symmetrical for all the indices. RTS takes the edge when it comes to risk-adjusted
returns. In most of the cases RTS, although volatile, performs better. The performance
measures indicate that the Russian market is inherently a risky market and therefore
caution must be taken from investment perspective.
7. Conclusions
In this paper, we examined the Russian Government argument that the shocks in US
markets are primarily responsible of increased volatility in the Russian equity markets
during September-October 2008. To neutralize the incoming shocks to Russian stock
market, the Russian Government decided to intervene and halt the stock trade.
We considered the case of RTS along with S&P500, DAX30, CAC40 and FTSE100
representing Russia, US, Germany, France and UK stock indices. In order to identify
the structural breaks, we employ MRSM. The results from MRSM suggest that
Russian Government intervention was mistimed and was ad hoc in nature.
Next we employed GARCH-DCC model to establish bivariate correlation between
RTS and the foreign markets. The results indicate weak linkages with the USA while
moderately strong (correlation varies between 0.4 and 0.7) integration with the
European equity markets. The result signi?es the effect of underlying ?nancial and
economic linkages between Russia and US and European stock markets. The study
concludes that the shocks that originated in the Russian market affected the foreign
markets while the volatility spillover from foreign markets affected the Russian
market. The bivariate GARCH-DCC analysis indicates that the government
intervention failed to calm the markets. In addition, we estimate GARCH-DCC model
for the entire sample by including and excluding RTS in order to evaluate the RTS
level of integration with the other markets. The results suggest that the S&P500,
DAX30, CAC40 and FTSE100 are strongly integrated (mean correlation around 0.6)
with each other in all periods. The relationship indicates interdependence between the
US and European stock markets. The correlation among the sample market plunges
dramatically once the RTS is included in the portfolio indicating weak integration
between the RTS and the other markets. Similar results were obtained using
Markowitz mean-variance ef?cient frontier. The RTS portfolio performance worsens
once the government intervenes, suggesting additional uncertainty in RTS due to
government ad hoc interventions. The Russian Government direct intervention in RTS
cannot be endorsed.
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Notes
1. The stock indices are: S&P500(USA), DAX30(Germany), CAC40(France), FTSE100(UK) and
RTS50(Russia). The case of Russia has beenselectedsolelydue towell documentedintervention.
2. Besides being affected by the subprime mortgage crisis, the analysts linked the RTS
downfall to other economic and political events which occurred prior to 15 September 2008
?nancial debacle.
Periods
b
RTS S&P500 DAX30 CAC40 FTSE100
Info ratio Crisis begin
(3 April
2001-10 July
2008)
0.06 (0.01) (0.01) (0.02) (0.02)
Tracking error 0.02 0.01 0.02 0.02 0.01
Jensen’s a
(RAR%)
(0.02) (1.64) (0.01) (1.25) (0.01) (1.35) (0.02) (1.48) (0.02) (1.53)
M
2
a (RAR%) 0.01 (0.90) 0.00 (0.30) 0.01 (0.73) 0.01 (0.63) 0.00 (0.40)
Info ratio Intervention
begin (3 April
2001-15
September
2008)
(0.31) 0.12 0.09 0.10 0.06
Tracking error 0.04 0.02 0.02 0.02 0.02
Jensen’s a
(RAR%)
(0.02) (1.06) (0.01) (0.57) (0.01) (0.97) (0.01) (1.05) (0.01) (0.99)
M
2
a (RAR%) 0.00 (0.02) 0.00 (0.13) 0.00 (0.03) 0.01 (0.17) 0.00 (0.08)
Info ratio Intervention
ends (3 April
2001-11
October 2008)
(0.02) 0.025 0.07 0.08 0.07
Tracking error 0.10 0.05 0.06 0.07 0.07
Jensen’s a
(RAR%)
(0.02) (0.53) (0.01) (0.11) (0.01) (0.39) (0.01) (0.52) (0.01) (0.63)
M
2
a (RAR%) 0.01 (0.50) 0.00 (0.83) 0.00 (0.69) 0.01 (0.46) 0.01 (0.61)
Info ratio Crisis ends (3
April 2001-16
September
2009)
0.02 (0.02) (0.02) (0.02) (0.01)
Tracking error 0.04 0.03 0.03 0.03 0.03
Jensen’s a
(RAR%)
0.00 (0.10) (0.00) (0.10) (0.00) (0.10) (0.00) (0.10) 0.00 (0.10)
M
2
a (RAR%) 0.00 (0.13) 0.00 (0.04) 0.00 (0.05) (0.00) (0.03) 0.00 (0.07)
Notes:
a
Sharp ratio was found to be negative in most of the cases and therefore it has been omitted
from the table;
b
the time period used for estimating the performance measure is as follows; the “crisis
begin” period starts from 3 April 2001 until the crisis begin 10 July 2008; the “intervention begin”
period starts from 3 April 2001 until the intervention begin 15 September 2008 and so on; RAR – risk
adjusted return under respective method; M
2
a – Modigliani and Modigliani method; benchmark
index – FTSE all world index; risk free rate – six months US treasuries bill rate (10 July 2008 (1.67
percent), 16 September 2008 (0.84 percent), 10 November 2008 (0.29 percent), 16 September 2009 (0.10
percent))
Source: www.treasury.gov
Table VII.
Investment performance
a
Government
intervention in
Russian bourse
337
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3. Trade halt is de?ned within the context of FFMS intervention. The FFMS directed the stock
exchange to stop the trade either for the whole day or within the trading time and sometimes
only after few minutes of start of trade.
4. Halting the trade means zero liquidity, i.e. the buyers and sellers cannot trade, therefore no
price discovery takes place.
5. European Commission (http://ec.europa.eu).
6. For instance in 2007 the bilateral trade between Russia and Germany, the UK, the USA,
France stood at $26 billion, $11 billion, $12 billion and $8 billion, respectively.
7. www.consensuseconomics.com
8. RTS is oldest and most modernize stock exchange in Russia. On the index diversi?cation,
RTS rigorously follows the EU directive on index diversi?cation: of?cial journal of the EU,
Commission Directive 2007/16/EC of 19 March 2007.
9. We have used longer time period to serve two purposes: (1) to overcome the convergence
problem in GARCH-DCC model; and (2) to study the evolution of dynamic correlation over
time.
10. www.adrbnymellon.com/
11. The pre-crisis period has been truncated in Figure 4, i.e. from 3 April 2001 until 1 June 2007
in order to illustrate clearly the crisis and intervention periods.
References
Boyer, B., Gibson, M. and Loretan, M. (1999), “Pitfalls in tests for changes in correlation”,
Federal Reserve Board International Finance Discussion Paper 597, Federal Reserve
Board, Washington, DC.
Corsetti, G., Pericoli, M. and Sbracia, M. (2002), “Some contagion, some interdependence:
more pitfalls in tests of ?nancial contagion”, CEPR Discussion Paper 3310, CEPR, London.
Dungey, M., Fry, R.A., Gonzalez-Hermosillo, B. and Martin, V.L. (2006), “International contagion
effects from the Russian crisis and the LTCM near-collapse”, Journal of Finance, Vol. 2
No. 1, pp. 1-27.
Edwards, S. (2000), “Contagion”, World Economy, Vol. 23, pp. 873-900.
Engle, R.F. (2002), “Dynamic conditional correlation – a simple class of multivariate GARCH”,
Journal of Business and Economics Statistics, Vol. 20 No. 3, pp. 339-50.
Forbes, K.J. and Rigobon, R. (2002), “No contagion, only interdependence: measuring stock
market comovements”, Journal of Finance, Vol. 57, pp. 2223-61.
Gelos, G. and Sahay, R. (2000), “Financial market spillover in transition economies”,
IMF Working Paper 00/71.
Hamilton, J.D. (1989), “A new approach to the economic analysis of non stationary time series
and the business cycle”, Econometrica, Vol. 57, pp. 357-84.
Harvey, C.R. (1995), “Predictable risk and return in emerging markets”, Review of Financial
Studies, Vol. 8 No. 3, pp. 773-816.
Jokipii, T. and Lucey, B. (2007), “Contagion and interdependence: measuring CEE banking sector
co-movements”, Economic Systems, Vol. 31 No. 1, pp. 71-96.
Kaminsky, G. and Reinhart, C. (1999), “The twin crises: the cause of banking and
balance-of-payments problems”, American Economic Review, Vol. 89, pp. 473-500.
Khan, S. and Batteau, P. (2011), “Should the government directly intervene in stock market
during a crisis?”, Quarterly Review of Economics and Finance, Vol. 51, pp. 350-9.
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
King, M.A. and Wadhwani, S. (1990), “Transmission of volatility between stock markets”,
Review of Financial Studies, Vol. 3 No. 1, pp. 5-33.
Loretan, M. and English, W. (2000), “Evaluating correlation breakdowns during periods
of market volatility”, in Bank for International Settlements (Ed.), International Financial
Markets and the Implication for Monetary and Financial Stability, BIS Conference Papers 8,
BIS, Basel, pp. 214-31.
Saleem, K. (2008), “International linkage of the Russian market and the Russian ?nancial crisis:
a multivariate GARCH analysis”, Research in International Business and Finance, Vol. 23,
pp. 243-56.
About the authors
Dr Salman Khan is PhD in Quantitative Finance and currently works as Assistant Professor at
Suleman Dawood School of Business, Lahore University of Management Sciences, Pakistan.
Salman Khan is the corresponding author and can be contacted at: [email protected]
Dr Pierre Batteau is PhD in Finance and presently works as Head of the Finance Department
at I.A.E Aix en Provence, Universite´ Aix-Marseille II, France.
Government
intervention in
Russian bourse
339
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_387656337.pdf
In September-October 2008 the Russian stock markets came under severe strain amidst
the global financial crisis. During this time the Russian government intervened several times to halt
the trade to impede the continuous slide. The government justified its actions owing to the argument
that the crisis was due to a trickledown effect from the financial crisis in the USA and the other
developed markets. The purpose of this paper is to put to test the government’s claim by exploring the
level of integration between Russia and the USA and European equity markets.
Journal of Financial Economic Policy
Government intervention in Russian bourse: a case of financial contagion
Salman Khan Pierre Batteau
Article information:
To cite this document:
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
Permanent link to this document:http://dx.doi.org/10.1108/17576381211279299
Downloaded on: 24 January 2016, At: 21:45 (PT)
References: this document contains references to 15 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 202 times since 2012*
Users who downloaded this article also downloaded:
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
Dimitrios Dimitriou, Theodore Simos, (2013),"Contagion channels of the USA subprime financial crisis:
Evidence from USA, EMU, China and J apan equity markets", J ournal of Financial Economic Policy, Vol. 5
Iss 1 pp. 61-71http://dx.doi.org/10.1108/17576381311317781
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
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Government intervention
in Russian bourse: a case
of ?nancial contagion
Salman Khan
Department of Finance, Suleman Dawood School of Business,
Lahore University of Management Sciences, Lahore, Pakistan, and
Pierre Batteau
Department of Finance, I.A.E Aix en Provence,
Universite´ Aix-Marseille II, Aix en Provence, France
Abstract
Purpose – In September-October 2008 the Russian stock markets came under severe strain amidst
the global ?nancial crisis. During this time the Russian government intervened several times to halt
the trade to impede the continuous slide. The government justi?ed its actions owing to the argument
that the crisis was due to a trickledown effect from the ?nancial crisis in the USA and the other
developed markets. The purpose of this paper is to put to test the government’s claim by exploring the
level of integration between Russia and the USA and European equity markets.
Design/methodology/approach – The study employs Markov Regime Switching Model for
tracking structural breaks in the time series. This method divides the data into three periods, i.e. pre
crisis, during crisis and post crisis. Next the Multivariate GARCH-DCC model technique is used to
establish the time varying linkages in order to verify the contagion effect. In the ?nal step the
Markowitz mean-variance framework is used to position each individual index portfolio with
respect to the ef?cient frontier to analyze the impact of crisis as well as Russian government
intervention.
Findings – The ?ndings suggest that the Russian equity market is weakly integrated with US and
strongly integrated with European markets. The results correspond to the underlying ?nancial and
economic linkages between Russia, the US and Europe. When examined in a portfolio setup, the results
show sudden fall in correlation among the Russian, US and European equity markets suggesting weak
linkages among these markets. Finally, the Markowitz Ef?cient frontier indicates dramatic rise in
volatility on the day intervention began and ended which signi?es the increased uncertainty among
the investors owing to Russian government ad hoc interventions.
Originality/value – The paper attempts to examine the Russian government intervention in the
backdrop of ?nancial crisis 2008 and concludes that the government intervention essentially increased
the uncertainty in the local as well as international markets. Therefore, it is essential that the
government should avoid direct intervention in its stock market.
Keywords Multivariate GARCH, Volatility spillovers, Financial crisis, Contagion, Stock markets,
Russia, United States of America, Europe
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?cation – C32, G15, G1
The authors wish to thank the Editor and two anonymous referees for constructive comments
on the earlier draft of this paper. The usual disclaimer applies.
JFEP
4,4
320
Journal of Financial Economic Policy
Vol. 4 No. 4, 2012
pp. 320-339
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381211279299
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
1. Introduction
The 15 September 2008 remains one of the most dramatic and memorable days in Wall
Street’s history when the Bank of America bought a near default Merrill Lynch for
$50 billion, Layman Brothers ?led bankruptcy and American International Group
(AIG) sought bailout from the US Government. On the 16 September 2008 the ?nancial
markets in the USA were in severe panic which triggered ?nancial crisis that
subsequently spread across the world markets.
For instance, the US, German, French, UK and Russian stock indices[1] recorded a
one day fall of 4.8, 2.8, 3.9, 4.0 and 5.0 percent, respectively, on 16 September 2008.
In the ensuing months, these stock markets continued to exhibit high levels of
volatility. Unlike US and European Governments, the Russian Government through
Federal Financial Market Service (FFMS) intervened in its stock market in order to
arrest the steep fall and suspended trade on 16 September 2009. The FFMS continued
to suspend the trade on different occasions spanning over 20 days; from 15 September
to 11 October 2008. The 20 days intervention appeared to be aligned with the steep fall
in US and European markets[2].
Throughout September-October 2008, the Russian Government claimed that the
shocks to Russian Trading System (RTS – a capital weighted index) directly resulted
from US subprime mortgage crisis and therefore necessitated government intervention.
The Russian Government held the viewthat trade -halt[3] should effectively isolate RTS
from external shocks and in turn provide time to investors for making informed
investment decisions (ef?cient market hypothesis-EMH). The Russian Government
adopted the strategy to halt the stock trade as soon as the markets in the USA and
Europe fell and it allowed the stock trade as soon as the US and European markets
exhibited the signs of recovery. Effectively, the government actions can be considered as
a hedging technique, i.e. to hedge the investors against market fall.
In this paper, we test the government claim that shocks from the ?nancial crisis
in 2008 transmitted from US and Europe to Russian stock markets, i.e. whether the
US and European stock markets can be considered contagious from Russian investor
perspective. In addition, we test the effectiveness of the government intervention in its
stock market in order to reduce the return-volatility and arrest the rapid decline
in stock prices.
At least in theory the notion of trade halt is acceptable since in the absence of trade the
price discovery[4] process does not take place. Such a theory would be more pronounced
in case of a single event such as the 9/11 terrorist attack on World Trade Center.
However, in the case of ?nancial crisis 2008, which spread over two months, this would
essentially mean that the government would halt the stock trade time and again
resultantly causing added uncertainty among the investors. In theory FFMS strategy
appears to be rewarding however, on the basis of risk-return, the Russian stock market
performed even worse than the US and European stock markets during the crisis.
2. Financial contagion
A market is regarded as ?nancially contagious if it spreads the shocks to other
markets. Edwards (2000) de?ne contagion as the transfer of shocks across countries
while Kaminsky and Reinhart (1999) de?ne contagion as the situation where the
knowledge of crisis in one country increases the risk of crisis in another country.
Forbes and Rigobon (2002) de?ne contagion as “a signi?cant increase in cross-market
Government
intervention in
Russian bourse
321
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
linkages after a shock to one country (or group of countries)” also called correlation
breakdown. Otherwise, a continued market correlation at high levels is considered to
be “no contagion, only interdependence”. When de?ning link between two or more
markets to be either contagious or interdependent is usually based on what triggers the
crisis in another market.
The most fundamental cause of contagion is considered to be a common shock,
for instance a major economic shift in industrial countries, a change in commodity prices,
a reduction in global growth, trade and ?nancial linkages ( Jokipii and Lucey, 2007).
Contagion can also be explained through the investor behavior theories such as the
liquidity problem. In this case investors cover their losses in one country by selling
securities inthe other markets inorder to raise cashinanticipationof greater redemptions.
Additionally, if banks experience amarkeddeteriorationinthe qualityof their loans toone
country, these banks may attempt to reduce the overall risk of their loan portfolio by
also minimizing their exposure in other high-risk investments, which could include other
emerging markets. Faced with liquidity problems, investors may be required to sell other
assets in their portfolios, ultimately leading to a fall in asset prices outside of the
crisis country, causing the disturbance to ripple through a variety of markets.
Much of the empirical work on measuring the existence of contagion is based on
comparing correlation coef?cients for interest rates, stock prices and sovereign spreads
between markets during a relatively stable period with a crisis or turbulent period
(King and Wadhwani, 1990; Boyer et al., 1999; Loretan and English, 2000; Forbes and
Rigobon, 2002; Corsetti et al., 2002).
We de?ne the contagion test given in Jokipii and Lucey (2007): if two markets
are naturally moderately correlated during periods of stability, then a shock to one
market will result in a signi?cant increase in market co-movements. This increase
constitutes contagion. If, on the other hand, the relationships do not change signi?cantly
after a shock to one market, and stability in the transmission mechanism persists, then
continued market co-movements can be inferred to as being driven by strong real
linkages betweenthe two economies. Suchstabilityinparameters over time woulddenote
interdependence. Based on these assumptions, contagion implies that cross-country
linkages are fundamentally different after a shock to one market while interdependence
implies no real change to relationships.
In establishing the relationship between the Russian and the foreign equity markets,
it is noteworthy that Gelos and Sahay (2000) and Saleem (2008) studied the 1998
Russian crisis and concluded in favor of contagion presence before the crisis in
the Russian market. Dungey et al. (2006) also found evidence of contagion from Russia
to international markets. These studies conclude that Russia is contagious from the
foreign markets perspective. More recently, Khan and Batteau (2011) evaluated the
performance of Russian Government intervention in RTS using event study
methodology. The study did not ?nd support in favor of government intervention.
However, the study is limited in its use of arbitrary linkages for cross-market volatility
transmission.
This paper attempt to ?ll this gap and attempt to answers following questions: does
Russia have ?nancial linkages with US and European stock markets? Are these
linkages strong enough to trigger contagion effect in the backdrop of ?nancial crisis
2008? Is there supporting evidence of underlying economic linkages which justi?es the
contagion effect?
JFEP
4,4
322
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In order to analyze the cross-country linkages most of the studies have focused on
multivariate volatility models. Consistent with these studies, we utilize the multivariate
generalized autoregressive conditional heteroscedasticity (GARCH) (dynamic
conditional correlation (DCC)) model. We establish DCC thresholds in order to clearly
understand and interpret the level of cross-market integration. Next we use Markowitz
ef?cient frontier framework to assess and verify the impact of Russian Government
intervention.
The rest of the paper is organized as follows: in Section 3 we highlight the economic
linkages underlying Russia and its counterparts, i.e. Europe and the USA. In Section 4
we discuss the data and preliminary analysis, in Section 5 we introduce GARCH-DCC
in bivariate and multivariate forms and the empirical results thereof, in Section 6 we
position each individual index portfolio with respect to the ef?cient frontier to analyze
the impact of crisis as well as Russian Government intervention and in Section 7 we
draw conclusions.
3. Economic linkages
The European Union (EU) and Russia have a strong trade relationship[5]. Bilateral
trade and investments continue to grow rapidly. Russia is one of the EU’s key trading
partners and trade between the two economies has showed steep growth rates until
mid-2008. This trend was interrupted by the economic crisis and the unilateral
measures adopted by Russia which have affected the bilateral trade[6]. The EU is by
far Russia’s main trading partner, accounting for 47.1 percent of its overall trade
turnover in 2010. It is also by far the most important investor in Russia. It is estimated
that up to 75 percent of foreign direct investment stocks in Russia come from the EU
Member States. EU good exports to Russia 2010 stands at $114.1 billion while EU
goods imports from Russia 2010 stands at $210.2 billion.
On the other hand, the level of bilateral ?nancial and economic exchange between
Russia and the USA is low[7]. Each only accounts for minor fractions of the other’s
trade with the rest of the world. Trade is dominated by inter-industry trade, with the
USA mainly exporting skill-intensive products to Russia while importing commodities.
This is not surprising given the countries’ different economic structures and factor
endowments. Russia accounted for only 0.7 percent of US exports and 1.3 percent of US
imports in 2008. Vice versa, the shares are slightly higher with the US accounting for
3.3 percent of Russia’s exports and 4.4 percent of Russia’s imports. Bilateral FDI shows
lower ?gures than trade: Russia accounted for only 0.3 percent ($6 billion) of the US
FDI stock abroad on average from 2000 to 2008. FDI increased sharply from 2003 to
2007 but recorded a signi?cant drop in 2008.
These economic linkages between Russia and Europe and the USA determine the
magnitude of shock transmitting from one market to another. Taking into account
these linkages, the correlation between Russia and Europe is expected to be stronger
and stable and the reverse is expected in case of Russia and the USA.
4. Data and preliminary analysis
Primary data consists of capital weighted indices including RTS 50 stock index[8]
(RTS), Standard and Poor’s 500 stock index (S&P500) – a US index, Deutscher Aktien
IndeX 30 (DAX30) – a German index, Cotation Assiste´e en Continu 40 stocks index
(CAC40) – a French index, Financial Times 100 stock index (FTSE100) – a UK index.
Government
intervention in
Russian bourse
323
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The data is obtained from DataStream comprising 2,266 daily observations from
3 April 2001 to 31 March 2010[9]. The closing prices of each index are converted to
natural log returns.
Apart from the underlying economic linkages as discussed in Section 3, the choice of
S&P500, DAX30, CAC40 and FTSE100 is due to the number of listed Russian
American Depository Receipts’ (ADR’s)[10] in these markets. Further support to
employ these series comes from the magnitude of trade speci?cally between Russia
and Germany, the UK, the USA and France. The area plot of the ?ve indices, i.e. RTS,
S&P500, DAX30, CAC40 and FTSE100 can be seen in Figure 1. The deep fall on
15 September 2008 has been attributed to the Lehman Brothers bankruptcy ?ling
among other events. The RTS appears more volatile from mid-2008 onwards.
As discussed earlier, the sudden fall on 15 September 2008 is quite steep compared to
the other stock indices. Since the time period under study includes a crisis period, we
need to identify appropriate breaks in the time series. This will allow us to analyze the
behavior of market correlation in different periods.
4.1 Structural breaks
Assume that y
t
is a time series generated as an autoregression of order r with regime
switching mean and variance:
y
t
¼ a
k
þb
k;S
t
x
k;t
þ1
t
ð1Þ
where:
1
t
, N 0; s
2
S
t
y
t
: RTS returns, a
k
: constant, b
k;S
t
: coef?cient for independent variables x
k,t
where
k ¼ 1, 2, 3, 4 representing S&P500, DAX30, CAC40 and FTSE100 returns, respectively;
1
t
: residual vector which follows a normal distribution; s
2
S
t
: the variance of the
Figure 1.
Daily closing prices of all
indices in the sample
7
Daily closing price
6
I
n
d
e
x
5
4
3
2
1
0
2002 2004 2006 2008 2010
RTS
S&P500
DAX
CAC40
FTSE100
JFEP
4,4
324
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
innovation at state S
t
and is assumed to be an n-state ?rst order Markov process,
taking the values 1, . . . ,n with transition probability matrix P ¼ P
ij
, i,j ¼ 1, . . . ,n
where:
P ¼ P
ij
½S
t
¼ jjS
t
¼ i? with
X
n
j¼1
P
ij
¼ 1 for all i:
Since S
t
is unobservable, the unknown parameters of the model can be estimated using
the non-linear ?lter proposed by Hamilton (1989). The model in equation (1) is
commonly known as Markov regime switching model (MRSM).
We estimate the model based on two states, i.e. low and high volatility for each pair
of index, i.e. RTS-S&P500, RTS-DAX30, RTS-CAC40 and RTS-FTSE100. The results
of MRSM are given in Table I.
For each pair of index, i.e. RTS-S&P500, RTS-DAX30, RTS-CAC40 and
RTS-FTSE100, we split the series into three periods based on two states:
(1) pre-crisis (3 April 2001-10 July 2008);
(2) crisis (11 July 2008-16 September 2009); and
(3) post-crisis (17 September 2009-31 March 2010).
Figure 2 shows the structural breaks in the sample markets. According to MRSM, the
crisis in RTS began on 11 July 2008, more than two months earlier than the beginning
of FFMS intervention. The FFMS intervened between 16 September 2008 and
11 October 2008 (20 days).
Figure 2 raises question regarding whether the timing of FFMS intervention
was consistent with the inception of crisis in RTS market. If the FFMS was of the view
that the RTS was suffering from the external shocks, then the FFMS should have
intervened in early July 2008. As a tertiary result, the Russian Government intervention
could be regarded as an ad hoc intervention.
4.2 Descriptive statistics
Table II gives the descriptive statistics for all the indices. These statistics has been
calculated using the daily closing returns of each index. The high return-volatility of
RTS in comparison to other indices is consistent with Harvey (1995) that emerging
stock markets show high volatility together with high returns. During pre-crisis,
the RTS offered high returns compared to the other stock markets while the riskiness
of RTS returns closely followed the other stock markets. The result shows that in
pre-crisis phase the RTS risk-return performance exceeded the other sample stock
markets.
During the crisis phase, all the markets exhibited negative returns along with high
levels of risks whereas RTS suffered the highest loss coupled with highest level of risk.
During this period, the risk-return of RTS was at extreme levels compared to the other
markets which points towards an abnormal activity that uniquely affected RTS
(possible the FFMS direct intervention). In post-crisis, RTS offered positive returns
with high risk compared to the other markets which essentially exhibited negative
returns with low level of risk. The argument shows the risk-return potential of RTS
market vis-a` -vis other markets.
Almost all the markets suffer from negative skewness and high kurtosis in all the
periods. During crisis the negative skewness shows that most of the returns lie under
Government
intervention in
Russian bourse
325
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
R
e
g
r
e
s
s
o
r
s
V
o
l
a
t
i
l
i
t
y
T
r
a
n
s
i
t
i
o
n
p
r
o
b
a
b
i
l
i
t
i
e
s
D
e
p
.
-
i
n
d
e
p
.
V
a
l
u
e
S
E
(
P
r
)
V
a
l
u
e
S
E
(
P
r
)
S
t
a
t
e
1
S
t
a
t
e
2
R
T
S
S
&
P
5
0
0
b
1
;
S
t
¼
1
0
.
3
3
*
0
.
0
4
(
0
.
0
0
)
s
21
;
t
0
.
0
0
0
2
*
0
.
0
0
(
0
.
0
0
)
p
1
1
¼
0
.
9
9
*
(
0
.
0
2
,
0
.
0
0
)
p
1
2
¼
0
.
0
6
*
(
0
.
0
1
,
0
.
0
0
)
b
1
;
S
t
¼
2
0
.
5
0
*
0
.
0
8
(
0
.
0
0
)
s
22
;
t
0
.
0
0
1
6
*
0
.
0
0
(
0
.
0
0
)
p
2
1
¼
0
.
0
1
*
(
0
.
0
0
,
0
.
0
0
)
p
2
2
¼
0
.
9
4
*
(
0
.
0
0
,
0
.
0
0
)
a
1
0
.
0
0
0
.
0
0
(
1
.
0
0
)
L
o
g
-
l
i
k
e
l
i
h
o
o
d
5
,
1
8
4
.
0
9
R
T
S
-
D
A
X
3
0
b
2
;
S
t
¼
1
0
.
3
4
*
0
.
0
2
(
0
.
0
0
)
s
21
;
t
0
.
0
0
0
2
*
0
.
0
0
(
0
.
0
0
)
p
1
1
¼
0
.
9
8
*
(
0
.
0
2
,
0
.
0
0
)
p
1
2
¼
0
.
0
6
*
(
0
.
0
2
,
0
.
0
0
)
b
2
;
S
t
¼
2
0
.
8
7
*
0
.
0
8
(
0
.
0
0
)
s
2
2
;
t
0
.
0
0
1
3
*
0
.
0
0
(
0
.
0
0
)
p
2
1
¼
0
.
0
2
*
(
0
.
0
0
,
0
.
0
0
)
p
2
2
¼
0
.
9
4
(
i
n
f
,
1
.
0
0
)
a
2
0
.
0
0
0
.
0
0
(
1
.
0
0
)
L
o
g
-
l
i
k
e
l
i
h
o
o
d
5
,
9
1
9
.
8
1
R
T
S
-
C
A
C
4
0
b
3
;
S
t
¼
1
0
.
4
0
*
0
.
0
3
(
0
.
0
0
)
s
21
;
t
0
.
0
0
0
2
*
0
.
0
0
(
0
.
0
0
)
p
1
1
¼
0
.
9
8
*
(
0
.
0
2
,
0
.
0
0
)
p
1
2
¼
0
.
0
6
*
(
0
.
0
1
,
0
.
0
0
)
b
3
;
S
t
¼
2
0
.
9
5
*
0
.
0
7
(
0
.
0
0
)
s
2
2
;
t
0
.
0
0
1
1
*
0
.
0
0
(
0
.
0
0
)
p
2
1
¼
0
.
0
2
*
(
0
.
0
0
,
0
.
0
0
)
p
2
2
¼
0
.
9
4
*
(
0
.
0
1
,
0
.
0
0
)
a
3
0
.
0
0
0
.
0
0
(
1
.
0
0
)
L
o
g
-
l
i
k
e
l
i
h
o
o
d
5
,
9
6
5
.
6
5
R
T
S
-
F
T
S
E
1
0
0
b
4
;
S
t
¼
1
0
.
5
1
*
0
.
0
3
(
0
.
0
0
)
s
21
;
t
0
.
0
0
0
2
*
0
.
0
0
(
0
.
0
0
)
p
1
1
¼
0
.
9
8
*
(
0
.
0
2
,
0
.
0
0
)
p
1
2
¼
0
.
0
5
*
(
0
.
0
1
,
0
.
0
0
)
b
4
;
S
t
¼
2
1
.
1
0
*
0
.
0
8
(
0
.
0
0
)
s
2
2
;
t
0
.
0
0
1
0
*
0
.
0
0
(
0
.
0
0
)
p
2
1
¼
0
.
0
2
*
(
0
.
0
0
,
0
.
0
0
)
p
2
2
¼
0
.
9
5
*
(
0
.
0
1
,
0
.
0
0
)
a
4
0
.
0
0
0
.
0
0
(
1
.
0
0
)
L
o
g
-
l
i
k
e
l
i
h
o
o
d
5
,
9
8
3
.
9
4
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
5
p
e
r
c
e
n
t
c
o
n
?
d
e
n
c
e
l
e
v
e
l
;
t
h
e
a
k
r
e
p
r
e
s
e
n
t
c
o
n
s
t
a
n
t
;
b
k
–
r
e
g
r
e
s
s
i
o
n
c
o
e
f
?
c
i
e
n
t
w
h
e
r
e
k
¼
1
(
S
&
P
5
0
0
)
,
2
(
D
A
X
3
0
)
,
3
(
C
A
C
4
0
)
a
n
d
4
(
F
T
S
E
1
0
0
)
;
s
2
1
;
t
–
v
a
r
i
a
n
c
e
o
f
t
h
e
i
n
n
o
v
a
t
i
o
n
a
t
s
t
a
t
e
S
t
w
h
e
r
e
S
t
¼
2
s
t
a
t
e
s
;
s
t
a
n
d
a
r
d
e
r
r
o
r
a
n
d
p
r
o
b
a
b
i
l
i
t
y
a
r
e
i
n
p
a
r
e
n
t
h
e
s
e
s
Table I.
Markov regime
switching model
JFEP
4,4
326
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the positive tail of the normal distribution indicating presence of extreme values
(the crisis) in left tail. The results from skewness and kurtosis are largely comparable
between RTS and the other stock markets which suggest RTS general behavior similar
to a stock market of a developed economy.
Figure 3 shows volatility clustering, i.e. large changes in returns are followed by
large changes while small changes are followed by small changes implying ARCH
effects. The strong ARCH effects can be found over the period ranging from July 2008
to September 2009 which is common to all the markets. This period is essentially
marked by the crisis phase. To deal with the ARCH effects, we use a parsimonious
model such as GARCH(1, 1)-DCC.
5. Dynamic conditional correlations
We use the following mean equation for estimation of each return series:
R
i;t
¼ a
i
þbR
i;t21
þ1
i;t
ð2Þ
where R
i,t
is the return on index i at time t, a
i
is a constant, b is the auto regression
coef?cient and 1
i,t
is the error term. The Engle (2002) GARCH-DCC model is a two-stage
estimator of conditional variances and correlations. In the ?rst stage, a univariate
GARCH model is estimated. In the second stage, the univariate variance estimates
obtained from the ?rst stage are used as inputs. The DCC model provides the time
Figure 2.
Estimated Markov regime
switching model
1
0.5
RTS-S&P500
RTS-DAX30
RTS-CAC40
RTS-FTSE100
0
1
0.5
0
1
0.5
S
m
o
o
t
h
P
r
o
b
a
b
i
l
i
t
i
e
s
0
1
0.5
0
2002 2004 2006 2008 2010
2002 2004 2006 2008 2010
2002 2004 2006 2008 2010
2002 2004 2006 2008 2010
State 1 State 2 Intervention begins Intervention ends
Government
intervention in
Russian bourse
327
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 3.
Daily return volatility of
all indices in the sample
0.4
–0.2
0.2
0.3
0.2
0.1
0
–0.1
0.3
0.2
0.1
0
–0.1
0
–0.4
0.3
0.2
0
0.1
2000
DAX
FTSE100 CAC40
S&P500
RTS
2005
2005 2000
2010 2000 2005 2010
2000 2005 2010 2000 2005 2010
2010
0.05
0
–0.05
–0.1
–0.1
0.1
Complete Pre-crisis Crisis Post-crisis
RTS Mean ( £ 1,000) 0.99 1.40 21.92 0.94
Median ( £ 1,000) 2.38 2.45 0.45 2.43
SD (%) 2.27 1.74 4.35 2.29
Skewness 20.49 20.54 20.16 20.51
Kurtosis 14.10 6.22 6.89 14.32
S&P500 Mean ( £ 1,000) 0.01 0.05 20.54 20.03
Median ( £ 1,000) 0.59 0.47 1.33 0.52
SD (%) 1.37 1.06 2.64 1.39
Skewness 20.10 0.09 20.08 20.08
Kurtosis 11.94 5.54 5.75 11.88
DAX30 Mean ( £ 1,000) 0.03 0.05 20.32 20.00
Median ( £ 1,000) 0.79 0.79 0.59 0.77
SD (%) 1.68 1.55 2.43 1.70
Skewness 0.06 20.12 0.37 0.06
Kurtosis 7.43 6.14 6.48 7.39
CAC40 Mean ( £ 1,000) 20.12 20.11 20.33 20.15
Median ( £ 1,000) 0.25 0.24 0.99 0.26
SD (%) 1.59 1.41 2.51 1.61
Skewness 0.06 20.05 0.20 0.07
Kurtosis 8.32 6.39 6.22 8.36
FTSE100 Mean ( £ 1,000) 0.00 20.02 20.15 20.04
Median ( £ 1,000) 0.47 0.39 0.47 0.39
SD (%) 1.34 1.15 2.27 1.36
Skewness 20.11 20.16 20.02 20.10
Kurtosis 9.80 6.42 6.48 9.81
Table II.
Descriptive statistics
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
varying correlations among the variables that indicates whether the co-movement of
the variables increase/decrease over time. This modeling technique is useful in
analyzing the level of market integration, a condition prerequisite for effective shock
transmission. Following Engle (2002), the DCC-GARCH model can be formulated
as follows:
r
t
jV
ðt21Þ
, Nð0; D
t
R
t
D
t
Þ ð3Þ
D
2
t
¼ diag{v
i
} þ diag{k
i
}r
t21
+r
t21
þ diag{l
i
}D
2
t21
ð4Þ
1
t
¼ D
21
t
r
t
ð5Þ
Q
t
¼ Sð1 2a 2bÞ þa 1
t21
1
0
t21
À Á
þbQ
t21
ð6Þ
R
t
¼ diag{Q
t
}
21
Q
t21
diag{Q
t
}
21
ð7Þ
where equation (5) represents the standardized errors, S is the unconditional
correlation matrix of the errors and W is the Hadamard product of two matrices of the
same size. The parameters of the DCC-GARCH model can be estimated using
maximum likelihood. If a þ b , 1 then equation (6) is mean reverting and the
log-likelihood for this estimator can be written as:
L ¼ 21=2
X
T
ðt¼1Þ
ðnlogð2pÞ þ 2 logðjD
t
jÞ þ logðjR
t
jÞ þ1
0
t
R
21
t
1
t
Þ
where D
t
¼ diag
??????
h
i;t
p È É
and R
t
is used to construct time varying correlations. We also
establish the rules for interpreting correlation between two or more series, i.e. strong
correlation: between 0.60 and 0.99, moderate correlation: between 0.40 and 0.59, weak
correlation: between 0.10 and 0.39.
5.1 Bivariate dynamic correlation
We estimate the GARCH(1, 1)-DCC model as laid down in equations (1)-(7). The results
are illustrated in Table III. The ARCH coef?cients a
RTS
and a
k
and the GARCH
coef?cients b
RTS
and b
k
are found to be signi?cant at the 5 percent con?dence level,
where k ¼ S&P500, DAX30, CAC40, FTSE100. The constant v
RTS
and v
k
are also
signi?cant at the 5 percent con?dence level. All the series show considerable level of
ARCH and GARCH effects.
The coef?cient a
m
is signi?cant at the 5 percent level, indicating the lingering effect
of standardized residuals in the previous period. The coef?cient b
n
is also signi?cant
at the 5 percent level, indicating the memory of correlations. For the entire sample
(a
m
þ b
n
) is close to 1, implying that the volatility is exhibiting a highly persistent
behavior. The high level of shocks (a
RTS
) originating in RTS indicate that RTS has
signi?cant role in in?uencing the correlation between RTS and the US and European
markets. On the other hand, the volatility spillover or GARCH levels (b
RTS
) are higher
in case of foreign markets compared with the RTS implying that the volatility spillover
in US and European markets largely affects correlation relationship with RTS. These
results signify that the RTS was less prone to the shocks and more to the volatility
spillover effects from the foreign equity markets. We plot the dynamic correlations
from Q
t
in Figure 4[11].
Government
intervention in
Russian bourse
329
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
All the paired correlations behave almost in a similar fashion. These begin sliding a
month before the crisis began, followed by a steep fall and then start rising.
The moderate correlation (#0.5) indicates that any setback in US or European
market does not have the potential to create a wider panic in RTS market and vice
versa, a result that supports the low level ARCH effects associated with these markets.
During the crisis, the correlation dips for a few months begin to move upwards during
the intervention period. The rectangular boxes in Figure 4 provide contagion build
RTS-S&P500 RTS-DAX30 RTS-CAC40 RTS-FTSE100
Value SE Value SE Value SE Value SE
v
RTS
0.12
*
0.00 0.12
*
0.00 0.12
*
0.00 0.12
*
0.00
a
RTS
0.10
*
0.00 0.11
*
0.00 0.11
*
0.00 0.11
*
0.00
b
RTS
0.87
*
0.00 0.86
*
0.00 0.86
*
0.00 0.86
*
0.00
v
k
0.01
*
0.00 0.02
*
0.00 0.02
*
0.00 0.01
*
0.00
a
k
0.07
*
0.00 0.09
*
0.00 0.09
*
0.00 0.09
*
0.00
b
k
0.93
*
0.00 0.90
*
0.00 0.90
*
0.00 0.90
*
0.00
a
m
0.01
*
0.00 0.02
*
0.00 0.02
*
0.00 0.02
*
0.00
b
n
0.99
*
0.00 0.97
*
0.00 0.97
*
0.00 0.97
*
0.00
Log
a
27,796.41 28,280.39 28,105.31 27,600.08
LBQ(i ) 19.01 (0.27)
*
17.45 (0.36)
*
18.89 (0.27)
*
16.24 (0.44)
*
LBQ(i )
2
16.53 (0.42)
*
13.46 (0.64)
*
13.84 (0.61)
*
16.16 (0.44)
*
LBQ(k) 31.24 (0.01) 18.96 (0.27)
*
29.36 (0.02) 25.67 (0.06)
*
LBQ(k)
2
18.78 (0.28)
*
166.15 (0.0) 230.34 (0.0) 545.46 (0.0)
Notes: Parameters signi?cant at:
*
5 percent con?dence level;
a
log likelihood function; LBQ – Ljung
Box Q test performed at 16th lag; LBQ(i ) – where i ¼ RTS; LBQ(k) – where k ¼ S&P500, DAX30,
CAC40, FTSE100
Table III.
GARCH-DCC (1, 1)
paired estimates
Figure 4.
Bivariate dynamic
conditional
correlation-GARCH(1,1)
1
0.5
RTS-S&P500
RTS-DAX30
RTS-CAC40
RTS-FTSE100
0
1
0.5
0
1
0.5
0
1
0.5
0
Contagion buildup
Contagion buildup
Contagion buildup
Contagion buildup
Contagion
Contagion
Contagion
Contagion
2008 2010 2009
2008 2010 2009
2008 2010 2009
2008 2010 2009
S
m
o
o
t
h
P
r
o
b
a
b
i
l
i
t
i
e
s
(
M
R
S
M
)
/
C
o
r
r
e
l
a
t
i
o
n
s
Intervention end Intervention begin Pre crisis crisis Post crisis Dynamic correlations
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
up phases and trigger the contagion process. The slow decay in correlations at the
beginning of the crisis indicates that the investors in foreign markets were liquidating
their positions. Once the views of the investors in RTS and foreign markets converge,
the correlations increases dramatically, i.e. contagion takes place. From government
intervention perspective, it is clear that halting the trade could not isolate RTS from the
global markets linkages. We see in Figure 4 that the bivariate correlation continues to
increase instead of decrease due to intervention. This provides evidence that the
halting the trade was not the solution even if the other markets are deemed as
contagious.
The descriptive statistics of correlations for the entire sample is given in Table IV.
The mean, maximum and standard deviation in correlations are much lower for
RTS-S&P500 (weakly correlated) compared to other paired indices indicating low level
of shock transmission and volatility spillover between these markets. In case of
RTS-DAX30, RTS-CAC40 and RTS-FTSE100, the correlations are moderate and suffer
from high volatility in pre-crisis, low volatility during crisis and post-crisis periods as
indicated by the standard deviation in Table IV. Here it is noteworthy that the Russian
Government pointing towards the USA as a contagious market is out of place since
the RTS correlation statistics are lower.
The reasons that explain as why RTS appears to be more correlated with the
European stock markets compared to US stock market lies in the underlying ?nancial
and economic linkages between these countries. As discussed earlier, Russia has
strong ?nancial and economic linkages with the Europe compared to the USA.
So far our results suggest that all the markets in the sample may be regarded as
contagious and the Russian Government intervention was not successful in reducing
the volatility in RTS market as well. To add further support to these interim ?ndings,
we conduct analysis by constructing portfolios that includes all sample indices with
and without RTS. If the inclusion of RTS increases the overall correlation of the
Parameters Pre-crisis Crisis Post-crisis
RTS-S&P500 Mean 0.22 0.30 0.43
Max. 0.37 0.39 0.49
Min. 0.03 0.11 0.35
SD 0.07 0.06 0.04
RTS-DAX30 Mean 0.35 0.47 0.56
Max. 0.67 0.61 0.65
Min. 20.02 0.19 0.44
SD 0.13 0.09 0.05
RTS-CAC40 Mean 0.39 0.50 0.61
Max. 0.68 0.66 0.68
Min. 20.01 0.18 0.48
SD 0.14 0.10 0.05
RTS-FTSE100 Mean 0.40 0.51 0.61
Max. 0.68 0.67 0.69
Min. 0.04 0.22 0.48
SD 0.13 0.09 0.05
Notes: Strong correlation: between 0.60 and 0.99; moderate correlation: between 0.40 and 0.59; weak
correlation: between 0.10 and 0.39
Table IV.
Descriptive statistics
of correlations
Government
intervention in
Russian bourse
331
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
portfolio then the Russian Government’s claim of shocks pouring from foreign markets
during crisis can be well endorsed. On the other hand if the inclusion of RTS decreases
the correlation among the global markets then the government claim could be
fairly rejected.
5.2 DCC-portfolio perspective
We estimate the GARCH(1, 1)-DCC model given in equations (1)-(7) for two different
group of portfolios:
(1) Portfolio-I consists of S&P500, DAX30, CAC40, FTSE100; and
(2) Portfolio-II consists of S&P500, DAX30, CAC40, FTSE100 and RTS.
For the ?rst portfolio, the results are given in Table V, Panel (a). The ARCH coef?cients
a
a
and the GARCH coef?cients b
a
are found to be signi?cant at the 5 percent
con?dence level. The shocks (a
a
) for DAX30, CAC40, and FTSE100 are relatively
higher than the shocks for S&P500 which indicate that the shocks originated in these
markets affects the USA more while the USA affects the European markets in terms of
volatility spillover (b
a
) effects. With regard to the persistence, the results for both
Panels (a) and (b) indicate that the sum of the estimated coef?cients of the variance
equation (a
m
þ b
n
) is close to unity. This implies that volatility exhibits a highly
persistent behavior in both groups.
For the second portfolio, the results are given in Table V, Panel (b). The
ARCHcoef?cients shocks (a
b
) and the volatility spillover (b
b
) are found to be signi?cant
at the 5 percent con?dence level. The shocks (a
b
) for RTS are high compared to other
markets which indicate that the shocks emanating from RTS are transmitted to these
markets. On the other hand, the volatility spillover (b
b
) is higher for the US and
RTS S&P500 DAX30 CAC40 FTSE100
Value SE Value SE Value SE Value SE Value SE
Panel (a): S&P500, DAX30, CAC40 and FTSE100
v
a
2.89
*
0.86 3.56
*
0.98 3.21
*
0.86 1.51
*
0.22
a
a
0.12
*
0.00 0.15
*
0.00 0.14
*
0.00 0.13
*
0.00
b
a
0.86
*
0.00 0.85
*
0.00 0.85
*
0.00 0.86
*
0.00
a
m
0.04
*
0.00 Log likelihood 230,641.45
b
n
0.94
*
0.00
LBQ(a) 40.40 (0.00) 28.04 (0.03) 25.85 (0.06)
*
17.13 (0.38)
*
LBQ(a)
2
16.66 (0.41)
*
10.81 (0.82)
*
10.99 (0.81)
*
14.16 (0.59)
*
Panel (b): RTS, S&P500, DAX30, CAC40 and FTSE100
v
b
0.11
*
0.00 0.01
*
0.00 0.02
*
0.00 0.01
*
0.00 0.01
*
0.00
a
b
0.10
*
0.00 0.07
*
0.00 0.09
*
0.00 0.08
*
0.00 0.09
*
0.00
b
b
0.87
*
0.00 0.92
*
0.00 0.91
*
0.00 0.92
*
0.00 0.91
*
0.00
a
m
0.02
*
0.00 Log likelihood 214,086.06
b
n
0.97
*
0.00
LBQ(b) 10.22 (0.85)
*
43.59 (0.00) 27.00 (0.04) 25.37 (0.06)
*
17.99 (0.32)
*
LBQ(b)
2
12.55 (0.71)
*
26.00 (0.05)
*
32.74 (0.01) 6.99 (0.97)
*
19.22 (0.26)
*
Notes: Parameters signi?cant at:
*
5 percent con?dence level; LBQ – Ljung Box Q test performed at
16th lag where a ¼ S&P500, DAX30, CAC40, FTSE100 and b ¼ RTS, S&P500, DAX30, CAC40,
FTSE100
Table V.
GARCH-DCC (1, 1)
group estimates
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
European markets compared to RTS. This suggest that the US and European markets
volatility spillover can upset the correlation with the RTS market, a result similar to
bivariate GARCH(1, 1)-DCC analysis. The comparative conditional correlation for the
two groups of samples is plotted in Figure 5 within MRSM setup.
The correlation for the Portfolio-I (i.e. excluding RTS) is strong with a mean greater
than 0.6 in all periods compared to the Portfolio-II (i.e. including RTS) where the mean
correlations are found to be less than 0.45. The descriptive statistics for the correlations
are given in Table VI. The high correlations in the Portfolio-I over all periods indicate
high level of market integration implying strong shock transmission among the
markets. Portfolio-II reveals low level of correlations and therefore implies low level
of shock transmission.
The US-Europe certainly enjoys a higher level of correlation (less volatile) which
indicates that the US and European markets are interdependent. As illustrated in
Table VI, the mean correlation reduces from 0.60 to 0.23 as RTS is included in the
Portfolio-II. This means that RTS changes the dynamics of relationship between US and
Figure 5.
Dynamic conditional
correlation-GARCH(1,1)-
portfolio setup
0.9
1
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
2009 2008 2010
Dynamic Correlation Comparison (Two Portfollos)
Interdependence
Contagion
Contagion Buildup
Pre crisis crisis Post crisis Portfolio (excI.RTS)
Intervention begin Intervention end Portfolio (incl.RTS)
Pre-crisis Crisis Post-crisis
Portfolio-I (excl. RTS) Mean 0.60 0.68 0.69
Max. 0.78 0.77 0.79
Min. 0.24 0.51 0.53
SD 0.09 0.06 0.06
Portfolio-II (incl. RTS) Mean 0.23 0.31 0.45
Max. 0.52 0.45 0.55
Min. 20.02 20.00 0.33
SD 0.09 0.09 0.07
Notes: Strong correlation: between 0.60 and 0.99; moderate correlation: between 0.40 and 0.59; weak
correlation: between 0.10 and 0.39
Table VI.
Descriptive statistics
of correlations
Government
intervention in
Russian bourse
333
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
European equity markets. The dramatic fall in correlation and rise in volatility
(in correlation) when switching from Portfolio-I to II suggests that the inclusion of RTS
contributes towards the overall risk of the portfolio.
The rise in correlation in Portfolio-II is quite subtle during the intervention period
compared to Portfolio-I. One can argue that the continuous rise during the intervention
period can be attributed to the Russian Government intervention, however, one need to
be cautious in interpreting this particular argument since we do not know what the
correlations would have been in the absence of government intervention.
The evidence suggest that RTS is weakly correlated with the global markets in
a portfolio setup implying that simultaneous shocks from these markets does not have
the potential to create wide spread crisis in RTS. However, the RTS does have the
potential to spillover volatility to these markets. This essentially means that RTS can be
regarded as a contagious market. In addition, similar argument can be used to justify
the portfolio diversi?cation bene?ts once the RTS is included in the portfolio. In the next
section we study the Portfolio-II in an ef?cient frontier framework with conditional risk
element. The analysis will allow us to understand the in?uence of each stock index
on the ef?cient frontier on crisis and intervention time periods.
6. Portfolio implication from intervention
In this section we evaluate the effects of intervention using portfolio theory. It would be
interesting to know the ef?cient frontiers and the performance of RTS vis-a` -vis other
indices including the respective position of each stock index. We construct ef?cient
frontier using mean-variance framework. We reutilize the conditional covariance
matrix estimated in the previous section for Portfolio-II which includes all the sample
indices and is shown in Figure 6.
The graph shows the time evolution of ef?cient frontier. It is quite vivid that in
2008-2009 all the world markets were going through the ?nancial crisis. During the
?nancial crisis the returns on ef?cient portfolios were as high as 50 percent per annum
and as low as 225 percent per annum. Similarly the risk ranged from a low level of
2 percent per annum to as high as 20 percent per annum. These re?ects the abnormal
Figure 6.
Mean-variance ef?cient
frontier (risk-return in
annualized percentage)
60
50
40
30
20
10
–10
–20
2000
2002
2004
2006
2008
2010
2012
Time Evolution of Efficient Frontier
0
5
10
15
20
25
Portfolio Risk
P
o
r
t
f
o
l
i
o
R
e
t
u
r
n
s
50
40
30
20
10
0
–10
0
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
times where even the ef?cient portfolios based on well diversi?ed sub portfolio such as
the sample indices are offering freighting risk-return. The subtleties of individual
index risk-return cannot be adjudged through a three dimensional ?gure. Therefore, it
is necessary that we analyze the performance of different indices in the portfolio at
different point of time using dynamic correlations and conditional volatilities. For this
purpose, we select four days in 2008-2009, i.e. the day the crisis begin, the days when
intervention begins and ends as well as on the day the crisis ends. Figure 7 shows these
four ef?cient frontiers. The risk in these ?gures re?ects the conditional variances
obtained from GARCH-DCC Portfolio-II estimation.
One can easily spot the behavior of RTS compared to other indices. RTS which
appear to be close to other indices on the day the crisis begin shored away from other
indices. Also, it is quite evident that RTS extreme position on the selected dates
in?uenced the ef?cient frontier to a great extent. Let us look at the individual index
positions in more detail.
As the crisis begins (11 July 2008) the RTS index portfolio return is better (1 percent
per annum) than the other indices except S&P500 (2 percent per annum). The positive
return on RTS comes with greater level of risk (12 percent per annum) compared
to other. On the intervention day (16 September 2008), the RTS return deteriorated to
230 percent per annum. The RTS risk pro?le dramatically changed as well on the
intervention day, i.e. from 12 percent per annum to 130 percent per annum. The risk
and return of the other index portfolios doubled from the crisis level, while witnessing
fall in returns except for S&P500 where the returns improved.
At the end of the intervention (11 October 2008), when the RTS was expected to
demonstrate that the government actions actually caused considerable reduction in the
risk, instead we ?nd that the risk increased to 275 percent per annum nevertheless the
returns improved by huge margin, i.e. 24 percent per annum compared to other indices.
This behavior of RTS appears consistent with the portfolio theory that high risk yields
Figure 7.
Mean-variance
ef?cient frontier
Government
intervention in
Russian bourse
335
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
high return. In this period, the other markets also suffered fromhigh volatility compared
to RTS and the returns in these markets improved as well. The large risk-gap between
RTS and the other index portfolios provide evidence that the intervention did not had
substantial effect on stabilizing the market conditions in RTS. The evidence con?rms to
earlier results that intervention created additional risk in RTS.
When the crisis ended on 17 September 2009, the RTS risk came down to 23 percent
per annum while the returns came down to 4.6 percent per annum. Table VII provides
key investment performance measures of these portfolios.
The tracking error is small for all the indices compared to RTS. The Jensen’s a is
negative for all the portfolios in all the periods’ except post-crisis day indicating
absence of positive alpha pro?ts for holding index portfolios. The M
2
a is positive and
symmetrical for all the indices. RTS takes the edge when it comes to risk-adjusted
returns. In most of the cases RTS, although volatile, performs better. The performance
measures indicate that the Russian market is inherently a risky market and therefore
caution must be taken from investment perspective.
7. Conclusions
In this paper, we examined the Russian Government argument that the shocks in US
markets are primarily responsible of increased volatility in the Russian equity markets
during September-October 2008. To neutralize the incoming shocks to Russian stock
market, the Russian Government decided to intervene and halt the stock trade.
We considered the case of RTS along with S&P500, DAX30, CAC40 and FTSE100
representing Russia, US, Germany, France and UK stock indices. In order to identify
the structural breaks, we employ MRSM. The results from MRSM suggest that
Russian Government intervention was mistimed and was ad hoc in nature.
Next we employed GARCH-DCC model to establish bivariate correlation between
RTS and the foreign markets. The results indicate weak linkages with the USA while
moderately strong (correlation varies between 0.4 and 0.7) integration with the
European equity markets. The result signi?es the effect of underlying ?nancial and
economic linkages between Russia and US and European stock markets. The study
concludes that the shocks that originated in the Russian market affected the foreign
markets while the volatility spillover from foreign markets affected the Russian
market. The bivariate GARCH-DCC analysis indicates that the government
intervention failed to calm the markets. In addition, we estimate GARCH-DCC model
for the entire sample by including and excluding RTS in order to evaluate the RTS
level of integration with the other markets. The results suggest that the S&P500,
DAX30, CAC40 and FTSE100 are strongly integrated (mean correlation around 0.6)
with each other in all periods. The relationship indicates interdependence between the
US and European stock markets. The correlation among the sample market plunges
dramatically once the RTS is included in the portfolio indicating weak integration
between the RTS and the other markets. Similar results were obtained using
Markowitz mean-variance ef?cient frontier. The RTS portfolio performance worsens
once the government intervenes, suggesting additional uncertainty in RTS due to
government ad hoc interventions. The Russian Government direct intervention in RTS
cannot be endorsed.
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Notes
1. The stock indices are: S&P500(USA), DAX30(Germany), CAC40(France), FTSE100(UK) and
RTS50(Russia). The case of Russia has beenselectedsolelydue towell documentedintervention.
2. Besides being affected by the subprime mortgage crisis, the analysts linked the RTS
downfall to other economic and political events which occurred prior to 15 September 2008
?nancial debacle.
Periods
b
RTS S&P500 DAX30 CAC40 FTSE100
Info ratio Crisis begin
(3 April
2001-10 July
2008)
0.06 (0.01) (0.01) (0.02) (0.02)
Tracking error 0.02 0.01 0.02 0.02 0.01
Jensen’s a
(RAR%)
(0.02) (1.64) (0.01) (1.25) (0.01) (1.35) (0.02) (1.48) (0.02) (1.53)
M
2
a (RAR%) 0.01 (0.90) 0.00 (0.30) 0.01 (0.73) 0.01 (0.63) 0.00 (0.40)
Info ratio Intervention
begin (3 April
2001-15
September
2008)
(0.31) 0.12 0.09 0.10 0.06
Tracking error 0.04 0.02 0.02 0.02 0.02
Jensen’s a
(RAR%)
(0.02) (1.06) (0.01) (0.57) (0.01) (0.97) (0.01) (1.05) (0.01) (0.99)
M
2
a (RAR%) 0.00 (0.02) 0.00 (0.13) 0.00 (0.03) 0.01 (0.17) 0.00 (0.08)
Info ratio Intervention
ends (3 April
2001-11
October 2008)
(0.02) 0.025 0.07 0.08 0.07
Tracking error 0.10 0.05 0.06 0.07 0.07
Jensen’s a
(RAR%)
(0.02) (0.53) (0.01) (0.11) (0.01) (0.39) (0.01) (0.52) (0.01) (0.63)
M
2
a (RAR%) 0.01 (0.50) 0.00 (0.83) 0.00 (0.69) 0.01 (0.46) 0.01 (0.61)
Info ratio Crisis ends (3
April 2001-16
September
2009)
0.02 (0.02) (0.02) (0.02) (0.01)
Tracking error 0.04 0.03 0.03 0.03 0.03
Jensen’s a
(RAR%)
0.00 (0.10) (0.00) (0.10) (0.00) (0.10) (0.00) (0.10) 0.00 (0.10)
M
2
a (RAR%) 0.00 (0.13) 0.00 (0.04) 0.00 (0.05) (0.00) (0.03) 0.00 (0.07)
Notes:
a
Sharp ratio was found to be negative in most of the cases and therefore it has been omitted
from the table;
b
the time period used for estimating the performance measure is as follows; the “crisis
begin” period starts from 3 April 2001 until the crisis begin 10 July 2008; the “intervention begin”
period starts from 3 April 2001 until the intervention begin 15 September 2008 and so on; RAR – risk
adjusted return under respective method; M
2
a – Modigliani and Modigliani method; benchmark
index – FTSE all world index; risk free rate – six months US treasuries bill rate (10 July 2008 (1.67
percent), 16 September 2008 (0.84 percent), 10 November 2008 (0.29 percent), 16 September 2009 (0.10
percent))
Source: www.treasury.gov
Table VII.
Investment performance
a
Government
intervention in
Russian bourse
337
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3. Trade halt is de?ned within the context of FFMS intervention. The FFMS directed the stock
exchange to stop the trade either for the whole day or within the trading time and sometimes
only after few minutes of start of trade.
4. Halting the trade means zero liquidity, i.e. the buyers and sellers cannot trade, therefore no
price discovery takes place.
5. European Commission (http://ec.europa.eu).
6. For instance in 2007 the bilateral trade between Russia and Germany, the UK, the USA,
France stood at $26 billion, $11 billion, $12 billion and $8 billion, respectively.
7. www.consensuseconomics.com
8. RTS is oldest and most modernize stock exchange in Russia. On the index diversi?cation,
RTS rigorously follows the EU directive on index diversi?cation: of?cial journal of the EU,
Commission Directive 2007/16/EC of 19 March 2007.
9. We have used longer time period to serve two purposes: (1) to overcome the convergence
problem in GARCH-DCC model; and (2) to study the evolution of dynamic correlation over
time.
10. www.adrbnymellon.com/
11. The pre-crisis period has been truncated in Figure 4, i.e. from 3 April 2001 until 1 June 2007
in order to illustrate clearly the crisis and intervention periods.
References
Boyer, B., Gibson, M. and Loretan, M. (1999), “Pitfalls in tests for changes in correlation”,
Federal Reserve Board International Finance Discussion Paper 597, Federal Reserve
Board, Washington, DC.
Corsetti, G., Pericoli, M. and Sbracia, M. (2002), “Some contagion, some interdependence:
more pitfalls in tests of ?nancial contagion”, CEPR Discussion Paper 3310, CEPR, London.
Dungey, M., Fry, R.A., Gonzalez-Hermosillo, B. and Martin, V.L. (2006), “International contagion
effects from the Russian crisis and the LTCM near-collapse”, Journal of Finance, Vol. 2
No. 1, pp. 1-27.
Edwards, S. (2000), “Contagion”, World Economy, Vol. 23, pp. 873-900.
Engle, R.F. (2002), “Dynamic conditional correlation – a simple class of multivariate GARCH”,
Journal of Business and Economics Statistics, Vol. 20 No. 3, pp. 339-50.
Forbes, K.J. and Rigobon, R. (2002), “No contagion, only interdependence: measuring stock
market comovements”, Journal of Finance, Vol. 57, pp. 2223-61.
Gelos, G. and Sahay, R. (2000), “Financial market spillover in transition economies”,
IMF Working Paper 00/71.
Hamilton, J.D. (1989), “A new approach to the economic analysis of non stationary time series
and the business cycle”, Econometrica, Vol. 57, pp. 357-84.
Harvey, C.R. (1995), “Predictable risk and return in emerging markets”, Review of Financial
Studies, Vol. 8 No. 3, pp. 773-816.
Jokipii, T. and Lucey, B. (2007), “Contagion and interdependence: measuring CEE banking sector
co-movements”, Economic Systems, Vol. 31 No. 1, pp. 71-96.
Kaminsky, G. and Reinhart, C. (1999), “The twin crises: the cause of banking and
balance-of-payments problems”, American Economic Review, Vol. 89, pp. 473-500.
Khan, S. and Batteau, P. (2011), “Should the government directly intervene in stock market
during a crisis?”, Quarterly Review of Economics and Finance, Vol. 51, pp. 350-9.
JFEP
4,4
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
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
King, M.A. and Wadhwani, S. (1990), “Transmission of volatility between stock markets”,
Review of Financial Studies, Vol. 3 No. 1, pp. 5-33.
Loretan, M. and English, W. (2000), “Evaluating correlation breakdowns during periods
of market volatility”, in Bank for International Settlements (Ed.), International Financial
Markets and the Implication for Monetary and Financial Stability, BIS Conference Papers 8,
BIS, Basel, pp. 214-31.
Saleem, K. (2008), “International linkage of the Russian market and the Russian ?nancial crisis:
a multivariate GARCH analysis”, Research in International Business and Finance, Vol. 23,
pp. 243-56.
About the authors
Dr Salman Khan is PhD in Quantitative Finance and currently works as Assistant Professor at
Suleman Dawood School of Business, Lahore University of Management Sciences, Pakistan.
Salman Khan is the corresponding author and can be contacted at: [email protected]
Dr Pierre Batteau is PhD in Finance and presently works as Head of the Finance Department
at I.A.E Aix en Provence, Universite´ Aix-Marseille II, France.
Government
intervention in
Russian bourse
339
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
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_387656337.pdf