investment positions data (using data from Lane and Milesi-Ferretti, 2006). The results
obtained prove to be very comparable with our results which indicates that focusing on
capital market data does not harm our analysis, but on the contrary is likely to achieve
more accurate results with a longer data coverage.
For further research it would be interesting to extend the country coverage to de-
veloping countries for whom economic smoothing might be even more crucial in light of
higher output volatility. Moreover, it would be interesting to know if international risk
sharing has increased over time and which role in this regard is played by the capital gains
channel. The role of financial deepening and home bias appears to be important as well.
Furthermore it is obvious that economic smoothing is only one part of international
investment decisions. Findings on gravity models of international asset trade prove to
be very significant (e.g. Aviat and Coeurdacier, 2006). Besides, the complete picture of
international portfolios also incorporates foreign assets and exchange rate considerations.
Obstfeld (2006) points out the importance of developing a consistent general equilibrium
portfolio-balance model. These models also have attracted a lot of attention recently, in
particular notably by Tille and van Wincoop (2007) as well as Devereux and Sutherland
(2007). It would be interesting to link their models to data on foreign assets and liabilities
in order to further evaluate the extent and potential of international risk sharing in times
of financial globalisation.
11