payments which represents a “benign loss” for the domestic economy. This decreases
domestic income and wealth commensurately, thus providing a smoothing or “hedging”
of the economic performance across the different states of the world.5 Obviously, this
smoothing mechanism also works when the economy performs poorly, since now there
should be capital losses (due to falling share prices) and lower income outflows.
This paper will examine if this mechanism is empirically observable which is “essen-
tial” (Obstfeld) in order to evaluate the stabilising effects of international investments.
Two main contributions are made: first the cyclicality of capital gains on equity and bond
markets is analysed on panel and country level; second, cross-country variation in cycli-
cality patterns is treated formally in order to find the fundamental reasons for differing
degrees of international risk sharing.6
This application is related to Davis, Nalewaik and Willen (2001) who develop a pro-
cedure to assess the gains to international financial trade in risky assets depending on
the correlations of domestic and international equity returns and domestic output inno-
vations.7
It is crucial to stress that the aim of this paper is not to provide an econometric model
that explains capital gains. But the emphasis rather is on the co-movement of capital
gains on different asset types and GDP growth in order to establish conclusions about
cyclicality and the associated international risk-sharing properties.
Accordingly, the rest of this paper is organised as follows: in the second section the
data will be presented. The empirical analysis starts in the third section by investigating
co-movements of domestic capital markets and GDP growth rates. Subsequently deter-
minants of country heterogeneity will be approached in section four; eventually some
concluding remarks will be made.
2 Data
In order to study the cyclical properties of capital markets we constructed a dataset of
20 industrial countries.8 This choice of the sample is very much determined by data
availability both in length and scope. We are able to capture the time series from 1973
5 If firms choose not to pay out dividends, but instead to keep retained earnings, the mechanism works
as well, since this should be reflected in higher stock prices and thus capital gains.
6This two-step approach is adapted from Lane’s (2003) cyclicality analysis on fiscal policy.
7 See their paper for a model of international trade in risky financial assets under incomplete markets.
8 Australia, Austria, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan,
Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom and the
United States.