these studies focus on the international diversification benefits from holding iShares funds of
different countries, in the short run and/or in the long run and compare their performance with the
respective closed-end country funds (CECF), when possible. In this backdrop, the present study
adds to the existing literature in a number of ways. First, it considers a very recent time period to
measure equity market integration relying on iShares prices using daily data from March 1996 to
January 2005. It is thus the longest time series of iShares prices analyzed in the literature so far.
Second, by drawing on the work of Olienyk, Schwebach and Zumwalt (1999), we perform a
more in-depth analysis of the long-term and short term interdependencies between iShares price
series than has been previously done using very recently developed time series techniques. In
particular, we use Gregory-Hansen (1996) test for cointegration that allows for an endogenously
determined structural break of unknown timing. We also investigate the time-varying dynamics
of the long-run relationships by using Hansen-Johansen (1999) recursive cointegration graphical
procedure. Furthermore, we use the recently developed dynamic conditional correlation
specification of multivariate GARCH models by Engle (2002) (DCC-GARCH) that allows for
explicit time variation in the conditional covariance and correlation matrix between iShares
returns.
Taking the U.S. based investor’s perspective, we measure the status of integration between
U.S. and the remaining Group of Seven (G7) markets: Canada, France, Germany, Italy, Japan and
United Kingdom. In our empirical analysis, we utilize prices for the iShares MSCI Index Series
for non-U.S. markets and Standard & Pours Depository Receipts (SPDRs) for the U.S. By using
G7 markets, we cover well in access of 60 % of world equity market capitalization (www.world-
exchanges.org/) and over 50 % of U.S. foreign equity holdings (US Department of Treasury
(2003)).
Based on our findings, there is some evidence in favor of multivariate cointegration