Handbook of Exchange Rates

14.5 Conclusion

Hansen and Hodrick (1980) and Fama (1984) show that the UIP is rejected by the data: changes in exchange rates do not correspond to interest rate differentials, and high interest rate currencies tend to appreciate and not depreciate as the UIP condition would suggest. This finding was considered for more than 20 years as proof of market inefficiency, not any more.

This survey shows that a recent stream of the exchange rate literature, based on new empirical findings, proposes new models of exchange rates. At the core of these models are novel SDFs. These risk-based approaches offer potential solutions to the UIP puzzle in purely rational and efficient asset markets. Rejection of the UIP puzzle is not proof of market inefficiency.

Risk-based models of exchange rates solve the UIP puzzle and open new directions for research in international finance. We mention here two potential avenues for research that build on these recent findings. First, more work is needed on the link between SDFs and economic fundamentals. Properties of exchange rates translate into necessary properties of SDFs. These SDFs should be heteroskedastic and they should exhibit some common components. Moreover, countries must be heterogenous: they should differ in their loadings on these common factors. It remains to link these properties to macroeconomic variables. What is, for example, the key source of heterogeneity across countries that will affect SDFs? Second, risk-based approaches imply interesting ...