4. Data and Methodology
4.1. Data
Our dataset consists of daily closing prices for SPDR for the U.S. market and six MSCI
iShares Series for the remaining G7 stock markets, namely U.K., Germany, France, Italy, Japan
and Canada. We concentrated on the iShares covering G7 countries as these represent the largest
equity markets, covering in excess of 85% of market capitalization of all world markets over the
period (www.world-exchanges.org/). The sample period spans almost ten years, from 18th March
1996 to 20th January 2005 and includes 2309 observations. The entire dataset is obtained from the
DataStream. The beginning of the sample period coincides with the start date of trading of
iShares, then WEBS. It is thus the longest time series of iShares prices analyzed in the literature
so far. Such a long dataset is especially suitable for applying cointegration analysis to
characterize the long term dependencies between the markets. Figure 1 below presents a graph of
SPDR and iShares prices over the sample period.
[Figure 1 around here]
Next we present descriptive statistics of the SPDR and iShares sample daily returns in
Table 2. The returns were calculated as continuously compounded returns, Rt = log Pt - log Pt-1 ,
where Pt is the closing price of a respective iShares. With the exception of Japan, even allowing
for the bear market of the early 2000’s, all securities displayed positive return. iShares’ returns
are leptokurtic for all G7 countries and mostly negatively skewed, except for Japan.
[Table 2 around here]
12