Despite widespread recognition of its major influence on open economies' macroeconomic behaviour, most theoretical models view the degree of international capital mobility as exogenous and typically perfect or zero while the available models of intermediate capital mobility treat the risk properties of foreign and domestic bonds as exogenous. In Discussion Paper No. 567, Research Fellow Neil Rankin develops a stochastic two-period model of a small open economy that derives the subjective probability distributions of real returns on bonds (which determine their substitutability in investors' portfolios), simultaneously with output and the exchange rate. Real returns on domestic bonds are uncertain because only the nominal interest rate is known in advance, while future inflation is unknown; while real returns on foreign bonds also entail uncertainty about the future exchange rate. Individuals base their decisions explicitly on utility maximization under uncertainty, and their subjective expectations of the probability distributions of returns are equated to the true probability distributions predicted by the model. Rational expectations therefore apply not only to the mean values of returns on bonds but also to their variability and correlation.

Rankin uses this model to assess the impact of an increase in uncertainty over future government policy, which is widely believed to raise the relative `risk premium' on domestic bonds. He finds that increased policy uncertainty may increase this premium, but this will depend on the source of uncertainty and the size of government relative to future output. If the uncertainty derives from the future money supply, there is no real risk premium, even though domestic and foreign bonds are imperfect substitutes. If it stems from government spending, however, government spending must be less than future output on average for the risk premium to rise rather than fall; and although this is the more likely case, it need not hold in an open economy. Rankin finds no simple relationship between the effects of policy uncertainty on the risk premium and current output: in the money supply case output falls despite the absence of a premium, while in the spending case output remains unchanged despite a rise in the premium.

Rankin cautions that these specific conclusions may not be widely applicable, since several strong assumptions are required to obtain a solution. Relaxing some of these particularly the small country assumption remains the subject of future work, but the effects on the risk premium and output of increased uncertainty concerning future policy are likely to continue to depend critically on its source.

Exchange Rate Risk and Imperfect Capital Mobility in an Optimizing Macromodel
Neil Rankin

Discussion Paper No. 567, July 1991 (IM)