Monetary and Portfolio Balance Models: Which does the Empirical Evidence Support?
In this chapter we seek to survey the empirical evidence on the two main classes of exchange rate models considered in previous chapters, namely the monetary (in both its flexible and sticky price versions) and the portfolio balance models. As has been demonstrated in previous chapters, the key distinguishing feature of these two classes of model concerns whether non-money assets (in particular government bonds) are perfect substitutes, and this may be summarised again with reference to equations (10.1) and (10.2),where (10.1) is a representation of risk adjusted UIP, and A denotes the risk premium which is determined in (10.2) by relative bond supplies (in this chapter all variables, apart from interest rates, are defined in natural logarithms). In the monetary class of models the bonds entering (10.2) are presumed perfect substitutes (that is, ϑ → 0) and therefore for this class of models λ t = 0 and (10.1) reduces to simple UIP:The whole motivation of the portfolio class of models is that bonds are, in fact, imperfect substitutes and therefore the appropriate way to define the relationship between domestic and foreign interest rates is given by (10.1). The issue of which of the two models is the most valid may therefore be seen to boil down to the issue of the existence of the risk premium, λ t . In this chapter we categorise the empirical exchange rate literature which has a bearing on this issue in the following way.
Unable to display preview. Download preview PDF.
© Emmanuel Pikoulakis 1995