Evolution of Inter and Intra-Regional Linkages to MENA Equity Market

1
Evolution of Inter and Intra-Regional Linkages to
MENA Equity Market

• Eric Girard
• and
• Eurico Ferreira

2
Introduction

• Mellon CapitalOur global tactical and strategic
  asset allocation productsinvest in developed
  country marketsand treat other less liquid
  markets as a block
• An important question examined in this paper is
  whether or not thin emerging capital markets such
  as MENA capital markets should be treated as a
  block in a global strategic or tactical
  portfolio.
• The proximity (geographic, culture, religion,
  etc.) of the countries may lead one to conclude
  that there is a close connection between their
  economies ? hence, susceptibility to be sensitive
  to shocks from neighboring countries.

3
MENA Markets some issues

• Wars, political turmoil, economic instability
  and institutional underdevelopment have
  traditionally been powerful obstacles to an
  increased access to MENA capital markets.
• Small capital markets with recent economic and
  financial development geared towards an increase
  in openness to foreign investors.
• During the nineties, Egypt, Israel, Jordan,
  Lebanon, Morocco, Tunisia and Turkey have been
  progressively lifting foreign investors
  ownership, and capital and dividends repatriation
  restrictions. Even the traditionally closed Gulf
  Country Council markets have become more
  accessible to foreign investors through
  international funds and trusts.

4
MENA Capital Markets overview (2000)
5
Financials MENA Vs. G7 and other EM countries
6
Lit. Review stuff

• Abraham, Seyyed and Al-Elg (2001) study Bahrain,
  Kuwait and Saudi Arabia using monthly index
  returns from 1993 to 1998, and observe low or
  negative correlations between markets?5 years, 3
  markets, 60 data points
• Omran and Gunduz (2001) use a multivariate
  cointegration methodology and find no long term
  stochastic trends between Jordan, Turkey, Egypt,
  Israel and Morocco from January 1996 to June
  1999?3.5 years, 5 markets, 42 data points
• Darrat, Elkhal and Hakim (2000) find long-term
  bivariate cointegrative relationships for
  Morocco-Egypt and Morocco-Jordan, but no
  multivariate cointegrative relationships between
  the three capital markets from October 1996 to
  August 1999 ?2.8 years, 5 markets, 34 data points

7
Motivation and research questions

• Previous studies? small samples, few markets?
  inexistence of intra-regional long-term
  (stochastic) price linkages.
• Remaining questions to be answered
• Any intra-regional spillovers (short-lived
  linkages)?
• Any inter-regional spillovers between MENA
  capital markets and other regional blocks?
• Any spillovers from the three major international
  financial crises that occurred during the 90s?
• Evolution of short-run price linkages that reveal
  a globalization trend as observed with most
  emerging markets during the 90s?

8
Data

• 11 MENA markets (Bahrain, Egypt, Israel, Jordan,
  Kuwait, Lebanon, Morocco, Oman, Saudi Arabia,
  Tunisia, and Turkey) and 5 regional indices
  (Asia AC, Europe, East Europe, Latin America AC,
  and North America)
• Daily, weekly and monthly frequency Index series
  (MSCI, IFC, Local? Datastream) Span 1990 to
  2001 also all data are in US Dollars?currency
  risk set to zero.
• Spillover study is done using daily data
• Capture potential short-lived interactionsI.e.cap
  ital movement are intrinsically short-term
  occurrences?Financial information networks are
  capable of disseminating news instantaneously
  around the world, a shock in a national stock
  market can be transmitted to another market
  within a very short period of time.
• Many of our series have less than eight years in
  coverage?serious methodological issues with using
  too few data points.
• Test results could be affected by infrequent
  trading.

9
Methodology

• 2 Market Linkage issues
• Global Strategic Asset Allocation (GSAA) Long
  horizon ?Cointegration analysis
• Series are I(1)??Tests of stationarity (ADF and
  KPSS)
• Bilateral cointgration (long term bivariate
  conintegrative relationship)
• Multilateral cointegration?Long-term common
  stochastic trends
• Global Tactical Asset Allocation (GTAA) Short
  Horizon ? Spillover Issue Pooled restricted
  GARCH-VAR methodology on price differences-I(0).
• Generalized Variance decomposition function
  (endogenous and exogenous of total variance
  forecast)?SIZE
• Generalized Impulse Response function forecast
  the effect of 1 SD shock in ALL endogenous
  variable? SIGN, TIMING
• Block Exogeneity Granger Causality?PREDICTABILITY
• Geweke Causality?INSTANTANEOUS SPILLOVER
• Dynamic of spillovers Pooled methodology

10
Long-term Stochastic Trend (we need to go through
this one before we do the fun stuff)

• Bilateral Cointegration Results
• Intra-regional long term linkages (4 out of 55)
• Bahrain-Jordan Israel-Turkey Morocco-Saudi
  Arabia Morocco-Tunisia
• Inter-regional long term linkages (6 out of 55)
• Israel-North America Morocco-North America
  Saudi-Arabia-North America Tunisia-North
  America Kuwait-East Europe Tunisia-Europe
• Multivariate cointegration analysis No common
  stochastic trends reverse cointegration tests
  (Hansen and Johansen, 1999) support findings
• Cointegration tests reveal some pairwise but no
  common stochastic trends to all MENA
  marketsi.e., no long-run co-movements

11
GARCH-VAR
This is a system, so variables are vectors

• Diagonal VAR-GARCH (see Engle and Sheppard,
  2001). By including a GARCH process for each
  equation of the VAR? heteroskedasticity is
  gonei.e., the greatest drawback of the VAR
  methodology.
• Covariances (as in an MGARCH model) are assume to
  be negligible so that our (Quasi) likelihood
  estimator will not die on usi.e., an MGARCH-VAR
  with 16 variables would require us to estimate,
  16 AR equations, 16 variance equations, and 120
  covariance equations.

12
Then, What? And How?

• Dynamic process
• ?VARGARCH is pooled forward
• multitude of GARCH-VARover time.
• multitude of GVDF, GIRF, GC and Geweke stuffover
  time.

• Amplitude?GVDF
• Sign?GIRF
• Timing?GIRF
• Direction
• Granger Causality
• (Lead-lag)
• Geweke Causality (contemporaneous)

13
Generalized Variance Decomposition Summary
Proportion of exogenous variance coming from MENA
Proportion of exogenous variance coming from
Blocks
14
GIR ?Intra-regional Exogenous Shocks Israel
15
GIR ? Intra-regional Exogenous Shocks Morocco
16
GIR ? Intra-regional Exogenous Shocks Jordan
17
GIR ? Inter-regional Exogenous Shocks Israel
18
GIR ? Inter-regional Exogenous Shocks Morocco
19
GIR ? Inter-regional Exogenous Shocks Jordan
20
Summary of shocks GIRF
The more integrated economies of Israel and
Turkey seem to process information flows from
global markets and act as conduits to other,
smaller, MENA markets.
Intra-Regional (Exogenous) Size Timing
Israel, Turkey Large High persistence
Morocco, Bahrain, Egypt, Kuwait, Lebanon, Oman, Tunisia Large Rapid decay
Jordan, Saudi Arabia Small Rapid decay
Inter-regional (Exogenous) Size Timing
Israel, Turkey Large Rapid Decay
Morocco, Bahrain, Egypt, Kuwait, Lebanon, Oman, Tunisia Small Rapid Decay
Jordan, Saudi Arabia Inexistent Rapid Decay
Endogenous shocks Size Timing
Israel, Turkey Large Rapid Decay
Morocco, Bahrain, Egypt, Kuwait, Lebanon, Oman, Tunisia, Jordan, Saudi Arabia Large High persistence
21
Causality Analysis

• Granger causality(lead-lag) Consistent with
  GVDF, No feed back relationships
• Geweke causality?contemporaneous relationship
• Most of the residuals correlation coefficients
  are positive ? investors view other regional
  economies as prone to different events.
• Residuals correlation are negative for four GCC
  market, indicating that capital tend to flow
  naturally from one market to another. Similar
  findings by Hassan (2003) who examines linkages
  among Bahrain, Kuwait and Oman stock markets from
  October 1994 and August 2001.
• Increase in contemporaneous spillover from other
  regional blocks
• Interesting case of Turkey and Israel? increasing
  amount of contemporaneous spillovers between
  Israel and the five regional indices? lead-lag
  spillover analysis fails to capture existing
  linkages that have become increasingly
  contemporaneous. The same conclusions can be
  drawn from Turkey and to a lesser extent for
  Morocco and Tunisia.

22
Conclusion

• Results from using the IR and VD functions
  illustrate a striking feature of MENA markets,
  namely the slow and small transmission of shocks
  during any period of this study. Results of our
  four spillover tests (Granger, GIRF, GVDF,
  residuals correlation, and Geweke) provide
  evidence that MENA markets are gradually opening
  to other regional and trans-continental
  economies, but remaining highly segmented (to the
  exception of Turkey and Israel) and perhaps
  predictable. In this case, tactical asset
  allocation strategies across MENA markets can be
  beneficial.