Standard models of international asset markets do not consider long-run output growth and typically estimate quite modest gains from asset trade, at least among industrial economies. Recent models of `endogenous growth' consider the effects of technological parameters, intertemporal preferences, market structures and government policies; in their extensions to multi- economy frameworks, international goods trade influences growth by shifting resources among alternative productive uses. In Discussion Paper No. 688, Research Fellow Maurice Obstfeld develops a simple model of global portfolio diversification that links growth to the opening of asset markets.

Each country can invest current resources in `safe' or `risky' projects; this captures the notion that continued growth depends on willingness to invest in the supply of specialized, and hence inherently risky, production inputs. Risky technologies have higher expected returns, but international asset trade allows each country to hold a globally diversified portfolio of risky investments, which encourages all countries to shift to high- return risky investments. Provided risky returns are imperfectly correlated across countries, and some risk-free assets are held initially, an increase in opportunities for diversification will always raise both expected growth and welfare.
For a closed economy, a reduction in uncertainty may spur economic growth. In open economies, all individuals will hold the same `global mutual fund' of risky assets, and global financial integration entails a world-wide portfolio shift into a risky `composite asset'. Obstfeld uses two-country examples to illustrate how certain structural assumptions lead to large welfare gains, but others result in the smaller gains found with exogenously determined long-run growth.

In an empirical example, partially calibrated to global consumption data, the estimated gains from moving to perfect global financial markets differ across regions and may be several times initial wealth. The assumptions about technologies and costs of capital relocation are rather unrealistic, however, and to replicate variations in countries' observed average rates of return to equity and real interest rates, this calibration exercise requires unrealistically high levels of individual risk aversion and intertemporal substitutability; it also assumes that all income risk can be traded. Obstfeld notes, however, that welfare gains of even a tenth of those this exercise predicts would be very significant; further research should therefore focus on developing simulation models rich enough to handle realistic complexity.

Risk-Taking, Global Diversification, and Growth
Maurice Obstfeld

Discussion Paper No. 688, August 1992 (IM)