ISSN 1836-8123
BRICs and PIGS in the presence of Uncle Sam and big
brothers: Who drive who? Evidence based on asymmetric
causality tests.
Abdulnasser Hatemi-J and Eduardo Roca
No. 2014-01
Series Editor: Dr. Alexandr Akimov
Copyright © 2014 by author(s). No part of this paper may be reproduced in any form, or stored
in a retrieval system, without prior permission of the author(s).
BRICs AND PIGS IN THE PRESENCE OF UNCLE SAM AND BIG
BROTHERS: WHO DRIVE WHO? EVIDENCE BASED ON
ASYMMETRIC CAUSALITY TESTS
Abdulnasser Hatemi-J
UAE University
E-mail: [email protected]
Eduardo Roca
Department of Accounting, Finance and Economics,
Griffith University
Nathan, Queensland, Australia 4111
Tel: +61-7-38757583, Fax: +61-7-3875 7760
Email: [email protected]
(corresponding author)
Abstract
We investigate the asymmetric causal interaction between the group of presently
depressed markets of Portugal, Italy, Greece and Spain (PIGS), on one hand, and the
booming markets of Brazil, Russia, India and China (BRIC), on the other hand, taking
into account Germany and France (Big Brothers) and the United States (Uncle Sam),
based on the asymmetric causality test methodology developed by Hatemi-J (2012).
We found that in both up and down market conditions, the BRICs influence the PIGS,
pulling the latter up during up market conditions and dragging them down during
down market times. The PIGS also affect the BRICs but only during down market
situations when the former pulls down the latter. Thus, the BRICs seem to be more
influential on the PIGS than the PIGS on the BRICs.
Keywords: BRIC, PIGS, asymmetric causality test, stock markets
JEL Classification: G15, C32
2
Introduction
The PIGS (Portugal, Ireland, Greece and Spain) are presently experiencing economic
difficulties while the BRICs (Brazil, Russia, India and China) are still relatively
booming. Recently, the BRICs have been asked to help rescue the PIGS as Big
Brothers (Germany and France) and Uncle Sam (the US) are dilly dallying. It may
not be well known but there has been a very strong economic and financial interaction
between the PIGS and the BRICs over the years. We therefore suspect that there has
been a significant linkage between the stock markets of these two groups of countries
and it would be therefore interesting to know as to who drive who. This paper
investigates the interdependence between the BRICs and the PIGS based on the newly
developed Hatemi-J (2012) asymmetric causality test. The test is very appropriate as
the situation of the two groups is asymmetric – the BRICs are emitting good news
while the BRICs, bad news. In order to avoid misspecification, we include the Big
Brothers and Uncle Sam as both the PIGS and the BRICs are highly linked to these
two.
There is now a voluminous literature on the issue of financial market integration.
However, there is no general agreement yet as to whether stock markets are integrated
or segmented. The findings of this literature are mixed depending on the set of
countries, time period and methodology being used. There is therefore still a need for
further studies on this issue. Our paper seeks to address this knowledge gap. To the
authors’ best knowledge, this is the first study that examines the stock market linkage
between the BRICs and the PIGS. As mentioned, it applies the asymmetric causality
test of Hatemi-J (2012). This newly developed causality test differs from other tests
which allow for asymmetric effects in that it explicitly separates the positive shocks
from negative shocks while the others utilize threshold, indicator variables or Markov
regimes.
Knowledge about the linkage between these two groups of important markets is one
that is very useful to investors and portfolio managers. This can assist them in
developing profitable trading strategies and in their quest for diversification of their
portfolios. For example, would it be a profitable strategy to short the PIGS and then
be long on the PIGS?
3
The results of this study could also be of assistance to
policymakers and regulators in terms of formulating policies that can contain
contagion risks. For instance, how can the BRICs ensure that the downturn in the
PIGS does not spill over to them? If it is found that there is a high linkage between
the PIGS and the BRICs, could this be a ground to convince the BRICs to assist the
PIGS in their recovery?
Literature on Financial Market Integration: An Overview
A huge number of studies have now investigated the issue of integration among stock,
bond, and money markets (e.g. Panopolou and Pantelidis, 2009; Chi, et al., 2006;
Click and Plummer, 2005; Roca, 1999). The ones pertaining to stock markets have
examined the following issues: (a) the extent to integration; (b) stability of linkages;
(c) groupings among markets in terms of linkages; and (d) the manner of interaction
among markets—which markets are influential, how one market affects another
market, and the speed of interaction among markets.
The results of stock market integration studies covering different countries, regions,
time periods and methods, however, have been mixed. For example, some find stock
markets to be integrated (e.g. Agmon, 1972; Ripley, 1973; Hillard, 1979; Ibbotson et
al., 1982; Jaffe and Westerfield, 1985; Schollhammer and Sand, 1987; Wheatley,
1988; Hamao, et al., 1990; Espitia and Santamaria, 1994, among others) while others
find them to be segmented (e.g. Grubel, 1968; Makridakis and Wheelwright, 1974;
Adler and Dumas, 1983; Jorion and Schwartz, 1986; Levy and Lerman, 1988; Dwyer
and Hafer, 1988; Jorion, 1989; Smith, et al., 1995). Similarly, some studies find
equity market linkages to be stable (e.g. Panton, et al., 1976; Philippatos, et al., 1983;
Goodhart, 1988) while others claim such linkages are unstable (e.g. Maldonado and
Saunders, 1981; Roll, 1989a and 1989b).
With regards to the structure of interaction among equity markets in terms of the
influence of one market on another, the manner of response of markets to influences
coming from other markets, and the speed by which shocks or volatility from one
market is transmitted to other markets, there is some evidence that the US might be
the world’s most influential stock market (e.g. Khoury, et al., 1987; Schollhammer
and Sand, 1987; Fischer and Palasvirta, 1990; Espitia and Santamaria, 1994).
However, the US market in turn might also be affected by other markets.
4
For
instance, in a study of the interaction among national equity markets, Huyghebaert
and Wang (2010) find that Singapore and Hong Kong Granger-cause the US market.
Further, some studies find no lead–lag relationships between markets (e.g. Granger
and Morgenstern, 1970; Hillard, 1979).
With respect to transmission of shocks
between markets, the results have also been mixed with some studies reporting that
the transmission process is efficient, i.e., occurring within a period of one to two days
(e.g. Schollhammer and Sand, 1987; Khoury, et al., 1987) while other studies (e.g.
Ng, et al., 1991) report the process to be inefficient.
Several studies point to the existence of linkages among certain groups of equity
markets based on some unifying or common factor, such as regional, economic, and
geographical relationships. To, et al. (1994) found the following clusters: Japan and
Asian emerging markets, and the UK and African emerging markets. Hillard (1979)
discovered a close association among intra–continental markets during the oil crisis of
1973 while Jorion (1989) reported a high degree of linkage among European
continental markets. An Anglo–Saxon cluster was also reported by Jorion (1989).
In light of the foregoing inconclusive evidence regarding integration and linkages—
that markets might or might not be integrated; that linkages might or might not be
stable; and that while the US stock market might be the most influential in the world,
it may itself be influenced by other markets—this study makes a significant
contribution to the literature by investigating, for the first time, the stock market
linkages between two very important blocs of markets – the BRICs and PIGS, which
are currently in the limelight and whose present conditions have provided lessons for
policymakers around the world. This is also the first stock market integration study
which applies the Hatemi-J (2012) asymmetric causality test.
The BRICs and PIGS and their Economic Interaction: A Brief Backgrounder
The acronym BRICs was coined by a Goldman Sachs investment banker, Jim O’Neill
in 2001 in a paper titled "Building Better Global Economic BRICs" 1. It refers to a
1
see
http://en.wikipedia.org/wiki/BRIC#cite_note-0#cite_note-0;
http://www.globaldashboard.org/2010/12/06/from-brics-to-pigs-whats-in-a-name/
5
country group that includes the four most promising emerging economies around the
world. These are Brazil, Russia, India, and China. As projected by O’Neill (2003), the
combined economies (in GDP terms) of Brazil, Russia, India and China would be
larger than those of G7 in 40 years, suggesting the significance of the BRICs as a
whole to the global economy. 2 As a matter of fact, the weight of the BRICs
economies is already 15% of the world economy in 2007, higher than the projection
of 10% in Goldman Sachs’s 2001 paper (O’Neill, 2007). At present, China and Russia
have accumulated trillions of dollars in foreign exchange reserves, and are now the
main creditors of western sovereigns. In the 1980s, emerging markets depended on
the west for capital inflows. Now, the situation is reversed, and the US and EU
depend on China to buy their sovereign debt 3. The BRICs possess a combined 4.3
trillion dollars in hard cash reserves, with China holding three-quarters of the kitty,
much of it in Euros 4. .
Given the global slowdown and the current eurozone debt crisis, a considerable
attention has been given to the BRICs with views that they can drive forward the
global economy. In fact, there has been proposals for the BRICs to come to the
rescue of a block of countries which are currently in dire economic situation. This
group is the so-called PIGS which refer to the countries of Portugal, Italy, Greece and
Spain. These countries have been victims of the recent Global Financial Crisis.
Before the crisis, these countries had high growth rates fuelled by massive spending
by both the government and private sectors with money flowing into speculative
sectors of their economies such as the real estate sector. When the GFC occurred and
liquidity vanished, their economies were hardest hit starting with their real estate and
banking sectors which then spread across other sectors including their governments.
These countries had to be bailed out with funds coming from the EU and the IMF
which happened only after so much dilly-dallying.
As mentioned, there have been calls for the BRICs to come to the rescue of the PIGS.
It is claimed that it is in the BRICs self-interest to undertake this, since although it is
2
G7 includes Germany, France, Italy, Japan, U.K., U.S., and Canada.
3
http://www.globaldashboard.org/2010/12/06/from-brics-to-pigs-whats-in-a-name/
4
http://www.cigionline.org/articles/2011/09/emerging-markets-hit-economic-stage-tonne-brics
6
not well known, there is actually significant economic linkage between these two
blocs of economies. There has been substantial trade between these two groups over
the years, with the BRICS enjoying surpluses consistently, as shown in Figure 1.
Billions
Figure 1. BRICs Exports to and Imports from PIGS
120
100
80
Exports
60
Imports
40
20
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
0
Source: World Trade Organisation
Given this substantial trade interaction between the BRICs and the PIGS, we expect
that there would also be significant linkage between the stock markets of these two
blocs of countries, with the BRICs granger-causing the PIGS. We therefore test this
expectation in this paper.
Methodology
We test the interaction between the BRICs and the PIGS controlling for the effect of
Uncle Sam (US) and Big Brothers (Germany and France). We collect weekly stock
market data for each of the PIGS, BRICs and Big Brothers, and also for Uncle Sam
during the said period. We take an average of the stock price of the BRICs to proxy
for the BRICs and we do the same for the PIGS. The empirical investigation in this
paper is based on the asymmetric causality test as suggested by Hatemi-J (2012). This
test separates the potential causal impact of positive shocks from the negative ones.
This is an important issue to take into account because people usually tend to react to
7
the negative news more than the good ones. This new methodology separates
explicitly the impact of permanent positive shocks from the negative ones unlike
other methods which allow for potential asymmetric impacts by using thresholds,
indicator variables or Markov-switching regimes.
In order to describe the asymmetric causality test in a simple way we concentrate on a
bivariate case. Consider the following two integrated variables, denoted by P1t and
P2t:
P1t = P1t −1 + ε1t = P1,0 +
t
∑ ε1i ,
i =1
(1)
and
P2t = P2t −1 + ε1t = P2,0 +
t
∑ ε 2i ,
i =1
(2)
where t =1,2,…T, the constants P1,0 and P2,0 are the initial values and ε1i and ε2i are
white noise error terms. Positive and negative changes of each variable are defined as
ε 1+i = max(ε 1i , 0 ), ε 2+i = max(ε 2i , 0), ε 1−i = min(ε 1i , 0) and ε 2−i = min (ε 2i , 0) , respectively.
This means that ε 1i = ε 1+i + ε 1−i and ε 2i = ε 2+i + ε 2−i . Based on this we can conclude that
P1t = P1t −1 + ε1t = P1,0 +
t
t
i =1
i =1
t
t
i =1
i =1
∑ ε1+i + ∑ ε1−i ,
and likewise
P2t = P2t −1 + ε1t = P2,0 +
∑ ε 2+i + ∑ ε 2−i
.
Thus, the positive and negative shocks of each variable in a cumulative format are
P1+t =
t
∑ ε1+i ,
i =1
P1−t =
t
∑ ε1−i ,
i =1
P2+t =
t
∑ ε 2+i
i =1
and P2−t =
t
∑ ε 2−i ,
respectively. These
i =1
cumulative components provide the possibility to implement asymmetric causality
8
between tests. For instance, in case the interest lies on testing for causality between
the positive components then the vector that should be used is Pt+ = ( P1+t , P2+t ) . This
vector can be used to estimate the following vector autoregressive model with the lag
order k, VAR (L): 5
Pt+ = ν + A1Pt+−1 +  + AL Pt+− k + ut+ ,
(3)
where v is the 2×1 vector of intercepts, and ut+ is representing a 2×1 vector of the
errors. Ar is a 2×2 matrix of parameters for lag order r (r = 1, …, k) to be estimated.
The optimal lag order k is selected by the minimization of the following information
criterion:
(
)
(
)

HJC = ln Ω f + k 2T −1 m 2lnT + 2m 2ln(lnT ) ,
k = 0, , k max .
(4)

here Ω f signifies the determinant of the variance-covariance matrix of the residuals
in the VAR model based on lag length k, m is the number of equations in the VAR
model, and T is the sample size. 6 After selecting the optimal lag order, the following
null hypothesis can be tested: 7
H0: the row j, column k element in Ar equals zero for r = 1,…, k.
5
(5)
For conducting tests for causality between negative cumulative components, the vector
Pt− = ( P1−t , P2−t ) is used.
6
The HJC is recommended in Hatemi-J (2003, 2008). Via Monte Carlo simulations the author showed
that this information criterion is successful in picking the correct lag order even if ARCH effects
prevail. The conducted simulations also showed that this information criterion has good forecasting
properties.
7
In order to account for the fact that each variable has a unit root, we have included an additional
unrestricted lag following the recommendations of Toda and Yamamoto (1995).
9
Before defining the Wald test we introduce some denotations to make the presentation
simple. Thus, we define the following: 8
(
Y : = P1+ ,, PT+
)
(m × T ) matrix,
D : = (v, A1 ,, Ak ) ( m × (1 + mk )) matrix,
 1 
 P+ 
 t 
+
Z t : =  Pt −1 


  
 +

 Pt − p +1 
((1 + mk ) × 1) matrix, for t = 1, …,T,
Z : = (Z 0 , , ZT −1 )
(
δ : = u1+ ,, uT+
)
((1 + mk ) × T ) matrix, and
(m × T ) matrix.
Via these denotations, the VAR(k) model can compactly be presented as
Y = DZ + δ ;
(6)
The Wald test to be used for the testing the null hypothesis of non-Granger causality,
H 0 : Rβ = 0 , is defined as
[(
)]
−1
Wald = (Rβ )′ R (Z ′Z )−1 ⊗ SU R′ (Rβ ) ,
(7)
where β = vec( D ) and vec is the column-stacking operator; ⊗ is the Kronecker
operator, and R is a k×m(1+mk) indicator matrix containing ones for restricted
parameters and zeros for the unrestricted parameters. The unrestricted variance-
δˆ ′δˆ
covariance matrix of the VAR model is SU = U U , where c is the number of
T −c
parameters in each of the equations in the underlying VAR. If the assumption of
8
It should be mentioned that initial values are assumed to be available. For more information on the
reason behind this assumption see Lutkepohl (2005).
10
normality holds, the Wald test in equation (7) is distributed as χ2 with k degrees of
freedom asymptotically. However, if the assumption of normality does not hold and if
ARCH effects exist then the Wald test will deviate for this assumed asymptotic
distribution. This is more likely to be the case when the sample size is small. A
potential solution to remedy this statistical problem is using the bootstrap simulations.
This procedure can be performed as follows. First run the regression model (6) with
the imposed restrictions under the null hypothesis of Granger non-causality. Next,
generate the bootstrap data, Yt ∗ , via Y * = Dˆ Z + δ * . It should be mentioned that the
bootstrapped residuals ( δ * ) are produced by T independent random draws from the
modified residuals of the regression. This bootstrap residuals are also mean-adjusted
in each simulation in order to make sure that the expected value of the residuals is
zero. The original residuals of the regression are also adjusted by using leverages in
order to achieve constant variance. 9 The bootstrapping is repeated 10000 times in
order to estimate the Wald test each time so as to create the distribution of the test.
Then, we take the (α)th upper quantile of the distribution of bootstrapped Wald test,
which is the bootstrap critical value at the α-level of significance, denoted by cα* .
Finally, we compare the calculated the Wald test, which is estimated by using the
original data, to the bootstrap critical value cα* . The null hypothesis can be rejected is
the estimated tests values is higher than the critical value. The bootstrap simulations
for implementing the asymmetric causality tests are performed via a statistical
software component written in GAUSS, which is available online. 10
RESULTS
Before testing for causality, we tested each variable for unit root. The results revealed
that each variable has indeed one unit root. We also tested the residuals of the VAR
model for multivariate normality and multivariate ARCH effects. The results showed
9
For additional information on leverage adjustment see Davison and Hinkley (1999) for univariate
analysis and Hacker and Hatemi-J (2006) for multivariate analysis.
10
See Hatemi-J (2011) for the statistical software component.
11
that the neither the assumption of normality nor the assumption of no ARCH can be
validated empirically. 11
The estimation results are shown in Table 1 in which we report the results pertaining
to the interaction between the BRICs and the PIGS.
Table 1. The Results of the Non-asymmetric Causality Tests
Null Hyothesis
Test
Bootstrap
Bootstrap
Bootstrap
Causal
Value
CV at 1%
CV at 5%
CV at 10%
Parameter
3.867
9.868
6.427
4.945
15.198*** 10.886
5.854
4.543
0.064
16.062*** 9.666
6.127
4.788
0.111
15.381*** 9.679
5.902
4.607
-0.174
1. The denotation A
B implies that variable A does not Granger cause variable
B.
2. CV is an abbreviation for the critical value.
3. An extra unrestricted lag was included in the VAR model in order to account
for the effect of a unit root.
It can be seen from the Table that a negative change in the PIGS causes positively a
negative change in the BRICs. This means that a downturn in the PIGS leads to a
greater downturn in the BRICs.
However, a positive change in the PIGS does not
affect the BRICs at all. Thus, the PIGS drag the BRICs down during a downturn but
does not pull it up during a market upturn.
A negative change in the BRICs granger-causes a negative change in the PIGS. This
means that the BRICs also pull down the PIGS during a market downturn. A positive
11
The results of these diagnostic tests are not presented here in order to save space. These results are
available on request.
12
change in the BRICs causes positively a positive change in the PIGS which means
that in an upturn, the BRICs pull up the PIGS, unlike the PIGS who do not carry the
BRICs up during a market upturn.
Hence, it appears that the BRICs are more
influential than the PIGS as they are able to affect the PIGS in both market conditions
while the PIGS only influence the BRICs during a downturn.
Conclusion
In this paper, we examine the interaction between the PIGs and BRICs based on the
asymmetric causality test of Hatemi-J (2012). Our results confirm the significant
interaction between the two blocs of markets. We find that the BRICs granger-cause
the PIGS in both up and down market conditions, driving the PIGS up when markets
are up and down when markets are down. The PIGs also granger-cause the BRICs
but this is only during down market conditions pulling down the BRICs when markets
are down. Thus, it appears that the BRICs are more influential on the PIGS than the
PIGS on the BRICs. These results reflect the economic interaction between the two
groups over the last decade in which the trade balance had been overwhelmingly in
favour of the BRICS. These findings also reaffirm the growing economic clout of the
BRICs which is now being reflected in financial markets. It also provides support to
the claim that the BRICs markets are also affected by conditions in the PIGS and
hence, there may be a case for the BRICs to take an active interest in helping ensure
the recovery of the PIGS for their own sake.
13
References
Adler, M., and Dumas, B. (1983) International portfolio choice and corporation
finance: a synthesis, Journal of Finance, 38, 925–984.
Agmon, T. (1972) The relations among equity markets: a study of share price comovements in the United States, United Kingdom, Germany and Japan, Journal of
Finance, 4, 839–855.
Calvi, R. (2010) Assessing financial integration: a comparison between Europe and
East Asia, Economic Papers, No. 423, European Commission.
Chi, J., Le, K., and Young, M. (2006). “Financial Integration in East Asian Equity
Markets”, Pacific Economic Review, 11(4), 513-526.
Claus E. and Lucey B. (2012) Equity Market Integration in the Asia Pacific Region:
Evidence from Discount Factors, Research in International Business and Finance,
forthcoming
Click, R.W. and Plummer, M.G. (2005). “Stock Market Integration in ASEAN after
the Asian Financial Crisis”, Journal of Asian Economics, 16, 5-28.
Davison, A.C. and Hinkley, D.V. (1999) Bootstrap Methods and Their Application,
Cambridge University Press, Cambridge
Doornik, J. A., & Hansen, H. (2008) An omnibus test for univariate and multivariate
normality, Oxford Bulletin of Economics and Statistics, 70, 927−939.
Dwyer, G., and R. Hafer, (1988), “Are National Stock Markets Linked?”, Review,
Federal Reserve Bank of St. Louis, 70, 3-14.
Espitia, M. and R. Santamaria (1994). “International Diversification Among the
Capital Markets of the EEC”, Applied Financial Economics, 4, 1, pp. 1-10.
Fischer, K., and A. Palasvirta (1990), “High Road to a Global Marketplace: The
International Transmission of Stock Market Fluctuations”, Financial Review, 25, 37194
14
Goodhart, C., (1988), “The International Transmission of Asset Price Volatility”,
Financial Market Volatility, Kansas City, MO: Federal Reserve Bank of Kansas City,
79-120.
Granger, C. W. J. (1969) Investigating causal relations by econometric models and
cross-spectral methods, Econometrica, 37, 424-439.
Granger, C.W.J., and O. Morgenstern, (1970), Predictability of Stock Market Prices,
Lexington, MA: Heath Lexington Books
Granger, C.W., Yoon, G. (2002) Hidden Cointegration. Department of Economics
Working Paper. University of California, San Diego.
Grubel, H.G., (1968), “Internationally Diversified Portfolios: Welfare Gains and
Capital Flows”, American Economic Review, 58, 1299-1314
Hacker R. S., and Hatemi-J, A., (2005) A multivariate test for ARCH effects, Applied
Economics Letters, Vol. 12, 411-417.
Hacker R. S., and Hatemi-J, A., (2006) Tests for causality between integrated
variables using asymptotic and bootstrap distributions: theory and application,
Applied Economics, Vol. 38, 1489-1500.
Hamao, Y., R.W. Masulis, and V. Ng, (1990), “Correlations in Price Changes and
Volatility across International Stock Markets”, The Review of Financial Studies, 3,
281-307.
Hatemi-J, A., (2003) A new method to choose optimal lag order in stable and unstable
VAR models, Applied Economics Letters, 10, 135-137.
Hatemi-J, A., (2011) ACTEST: GAUSS module to Apply Asymmetric Causality
Tests, Statistical Software Components, No. G00012, Boston College Department of
Economics. Link: http://ideas.repec.org/c/boc/bocode/g00014.html
Hatemi-J, A., (2012) Asymmetric tests for causality with an application, Empirical
Economics, Vol. 43, 447-456.
15
Hatemi-J A and Roca E (2004) An Examination of the Equity Market Price Linkage
between Australia and the European Union Using Leveraged Bootstrap Method,
European Journal of Finance, 10(6), 475-488.
Hillard, J. E. (1979), “The Relationship between Equity Indices on World
Exchanges”, Journal of Finance, 34, 103-114.
Huyghebaert, N. and L. Wang, (2010), “The Co-Movement of Stock Markets in Asia:
Did the 1997-1998 Asian Financial Crisis Really Strengthen Stock Market
Integration?” China Economic Review, 21, 98–112
Ibbotson, R.C., R.C. Karr, and A.W. Robinson (1982), “International Equity and
Bond Returns”, Financial Analysts Journal, 38, 61-83.
Jaffe, J. and R. Westerfield (1985), “Patterns in Japanese Common Stock Returns:
Day of the Week and Turn-of-the-Year Effects”, Journal of Financial and
Quantitative Analysis, 20, 261-272.
Jorion, P., (1989), “The Linkages between National Stock Markets”, in aliber, R.Z.
(ed.), The Handbook of International Financial Management, Illinois: Dow Jones –
Irwin
Jorion, P., and E. Schwartz (1986), “Integration vs. Segmentation in the Canadian
Stock Market”, Journal of Finance, 41, 603-614.
Khoury, S.J., B. Dodin, and H. Takada, (1987), “Multiple Time Series Analysis of
National Stock Markets and Their Structure: Some Implications, in Khoury, S.J. and
A. Ghosh (eds.), Recent Developments in International Banking and Finance, D.C.
Heath and Co.
Levy, H., And Z. Lerman, (1988), “The Benefits of International Diversification in
Bonds”, Financial Analysts’ Journal, 56-64
Makridakis,
S.G.,
and
S.C. Wheelwright,
(1974),
“An
Analysis
of the
Interrelationships among World Stock Exchanges”, Journal of Business Finance &
Accounting, 1, 195-215.
16
Maldonado, R., and A. Saunders, (1981), “International Portfolio Diversification and
the Inter-temporal Stability of International Stock Market Relationships, 1957-1978”,
Financial Management, 10, 54-63.
Ng, V., R.P. Chang, and R.W. Chou, (1991), “An Examination of the Behaviour of
Pacific Basin Stock Market Volatility”, in Rhee, S.G. and R.P. Chang (eds.), Pacific
Basin Capital Markets Research, 2, Holland: Elsevier Science Publishers, 245-260
Panopoulou, E., & Pantelidis, T., (2009), “Integration at a Cost: Evidence from
Volatility Impulse Response Functions”, Applied Financial Economics, 19(11), 9-17.
Ripley, D., (1973), “Systematic Elements in the Linkage of National Stock Market
Indices”, Review of Economics and Statistics, 55, 356-361.
Panton, D.B., V.P. Lessig, and O.M. Joy (1976), “Co-movement of International
Equity Markets: A Taxonomic Approach”, Journal of Financial and Quantitative
Analysis, 11, 415-32.
Philippatos, G.C., A. Christofi, and P. Christofi, (1983), “The Intertemporal Stability
of International Stock Market Relationships: Another View”, Financial Management,
12, 63-69.
Roca, E.D., (1999), “Short-term and Long-term Price Linkages between the Equity
Markets of Australia and its Major Trading Partners”, Applied Financial Economics,
9, 501-511.
Roll, R. (1989a), “The International Crash of October 1987”, in Kamphuis, R., R.
Karmendi, and J. Watson (eds.), Black Monday and the Future of Financial Markets,
Homewood, IL: Irwin, 35-70.
Roll, R., (1989b), “Price Volatility, International Market Links, and Their
Implications for Regulatory Policies”, Journal of Financial Services Research, 3,
211-246.
Schollhammer, H., and O.C. Sand, (1987), “Lead-Lag Relationships among National
Equity Markets: An Empirical Investigation”, in Khoury, S.J. and A. Ghosh (Eds.).
17
Recent Developments in International Banking and Finance, 1, Cambridge, MA:
Lexington, 149-168.
Smith, K. L., J. Brocato, and J.E. Rogers, (1995), “An Analysis of World Capital
Market Return/Risk Ratios: A Test of Global Financial Integration during the 19801991 Period”, Managerial Finance, 21, 13-31.
To, M.C., K.G. Assoe, and S. Pariente, (1994), “The Efficiency of Emerging Stock
Markets and Their Relationships with the World’s Major Stock Exchanges”,
Montreal: CETAI University of Montreal.
Toda, H.Y. and Yamamoto, T. (1995) Statistical inference in vector autoregressions
with possibly integrated processes, Journal of Econometrics, 66, pp. 225-50.
Wheatley, S., (1988), “Some Tests of International Equity Market Integration”,
Journal of Financial Economics, 21, 177-212
18