In this section, we consider the hypothesis that the negative correlation
between stock price co-movement and per capita income is due to "rms in
low-income economies having more correlated economic fundamentals. To test
this hypothesis, we include speci"c structural variables in the vector x
j that
might provide separate proxies for such an effect. These variables are macroeconomic
volatility, country size, and economy diversification. Since these variables
may not encompass all sources of market-wide price movement, we also include
a direct measure of earnings co-movement for firms in each economy using
standardize "rm-level accounting data. If including these variables in the
vector x
j renders per capita income insignificant in regression (11), we can
conclude that per capita income provides a proxy for these structural effects.
A description of each structural independent variables follows.