minor concern in our analysis.
Furthermore it is very crucial to know if extended periods of economic growth are
reflected in higher cumulative capital gains on the stock market. For this purpose we
constructed five-year GDP growth rates and cumulative five-year rates of stock market
capital gains. Since this evidently leaves fewer data points available, this question is only
analysed in panel format (with and without country fixed effects).
EQKG5it = αi + βg5it + uit (4)
where EQKG5 is the cumulative five year real rate of capital gains on the domestic stock
market index and g5 is the cumulative real rate of domestic GDP growth over five years.
The empirical evidence is very striking (Table 3): In terms of domestic currency the
coefficient is 7.58, in US dollar terms even 6.46 (both significant at the 1% level). When
estimating without country fixed effects, the coefficients are also significant at the 1%
level, but smaller in magnitude (4.95 and 4.39).9
Thus, there is strong pro-cyclical co-movement of domestic GDP growth and the stock
market over a five year horizon. This points towards domestic equity being “a claim on
GDP” possibly not in the short run (that is one year), but definitely in the medium run
of five years. Hence, in this time framework the necessary cyclical properties of the stock
market are satisfied in order to generate economic or wealth stabilisation as described
above. This result is very appealing as it offers risk sharing potential on a global scale in
particular when investments are made over a medium term horizon. Thus, equity capital
gains can act as an effective risk sharing device, when the investment behaviour reaches
the appropriate time frame. This result is in line with Giannone and Reichlin (2006) who
find increasing risk sharing particularly over long horizons.
Overall we found in this section that there is reasonable evidence on co-movement of
GDP growth and stock markets. In particular over a horizon of five years pro-cyclicality
of capital gains on equity is observable. This is driven by countries such as Australia,
Canada and the Netherlands. However, one has to stress that some countries exhibit
counter-cyclicality (in particular with regard to the idiosyncratic part of the domestic
stock market), whereas other show no cyclical movement at all.
9 The use of time fixed effects does not affect results in this analysis.